simple_knn.ipynb 64.5 KB
Newer Older
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import librosa\n",
    "import librosa.display\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import IPython\n",
    "import urllib\n",
    "from keras import models\n",
    "from keras import layers\n",
    "import seaborn as sns\n",
    "import os\n",
    "import tensorflow as tf\n",
    "from audio_classification.preprocess import preprocess\n",
    "from sklearn import preprocessing\n",
    "from keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "from sklearn.model_selection import train_test_split\n",
    "from audio_classification.dnn_simple_classifier.dnn_simple_classifier import DNNSimpleClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUMBER_OF_MFCCS = 13\n",
    "CLIP_SIZE = 1290\n",
    "SAMPLE_SHAPE = (NUMBER_OF_MFCCS, CLIP_SIZE, 1)\n",
    "MODEL_PATH = DNNSimpleClassifier.MODEL_PATH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cached features found\n"
     ]
    }
   ],
   "source": [
    "data, similarity = preprocess(NUMBER_OF_MFCCS)\n",
    "\n",
    "le = preprocessing.LabelEncoder()\n",
    "transformed = le.fit_transform(data['label'])\n",
    "l = []\n",
    "for index, row in data.iterrows():\n",
    "    arr = np.load(row['file'])\n",
    "    l.append(arr[:, :CLIP_SIZE])\n",
    "    \n",
    "X = np.expand_dims(np.stack(l), axis=3)\n",
    "y = np.array(transformed)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, stratify=y, random_state=666)\n",
    "X_train, X_valid, y_train, y_valid = train_test_split(X_train, y_train, stratify=y_train, random_state=666)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_loss_accuracy(history):\n",
    "    historydf = pd.DataFrame(history.history, index=history.epoch)\n",
    "    plt.figure(figsize=(8, 6))\n",
    "    historydf.plot(ylim=(0, max(1, historydf.values.max())))\n",
    "    loss = history.history['loss'][-1]\n",
    "    acc = history.history['accuracy'][-1]\n",
    "    plt.title('Loss: %.3f, Accuracy: %.3f' % (loss, acc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "average_pooling2d_1 (Average (None, 13, 1, 1)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 13)                0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 256)               3584      \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 32)                4128      \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 10)                330       \n",
      "=================================================================\n",
      "Total params: 40,938\n",
      "Trainable params: 40,938\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = models.Sequential()\n",
    "\n",
    "model.add(layers.Input(shape=SAMPLE_SHAPE, name='input'))\n",
    "model.add(layers.AveragePooling2D(pool_size=(1, CLIP_SIZE)))\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.2))\n",
    "model.add(layers.Dense(128, activation='relu'))\n",
    "model.add(layers.Dropout(0.2))\n",
    "model.add(layers.Dense(32, activation='relu'))\n",
    "model.add(layers.Dense(10, activation='softmax'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',\n",
    "              loss='sparse_categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 7.4633 - accuracy: 0.1300\n",
      "Epoch 00001: val_accuracy improved from -inf to 0.13500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 7.4633 - accuracy: 0.1300 - val_loss: 2.4958 - val_accuracy: 0.1350\n",
      "Epoch 2/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 2.9667 - accuracy: 0.1714\n",
      "Epoch 00002: val_accuracy improved from 0.13500 to 0.17500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.9235 - accuracy: 0.1717 - val_loss: 2.2705 - val_accuracy: 0.1750\n",
      "Epoch 3/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 2.5226 - accuracy: 0.1837\n",
      "Epoch 00003: val_accuracy improved from 0.17500 to 0.24500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.4810 - accuracy: 0.1900 - val_loss: 2.1913 - val_accuracy: 0.2450\n",
      "Epoch 4/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 2.3742 - accuracy: 0.2064\n",
      "Epoch 00004: val_accuracy did not improve from 0.24500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.3586 - accuracy: 0.2100 - val_loss: 2.2201 - val_accuracy: 0.2000\n",
      "Epoch 5/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 2.2119 - accuracy: 0.2125\n",
      "Epoch 00005: val_accuracy did not improve from 0.24500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.2140 - accuracy: 0.2183 - val_loss: 2.1610 - val_accuracy: 0.2350\n",
      "Epoch 6/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 2.1622 - accuracy: 0.2500\n",
      "Epoch 00006: val_accuracy improved from 0.24500 to 0.25000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 1s 14ms/step - loss: 2.1641 - accuracy: 0.2517 - val_loss: 2.0823 - val_accuracy: 0.2500\n",
      "Epoch 7/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 2.1578 - accuracy: 0.2886\n",
      "Epoch 00007: val_accuracy improved from 0.25000 to 0.30500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.1667 - accuracy: 0.2900 - val_loss: 2.0183 - val_accuracy: 0.3050\n",
      "Epoch 8/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 2.0125 - accuracy: 0.2910\n",
      "Epoch 00008: val_accuracy did not improve from 0.30500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 2.0248 - accuracy: 0.2983 - val_loss: 2.0300 - val_accuracy: 0.3050\n",
      "Epoch 9/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 2.0740 - accuracy: 0.2905\n",
      "Epoch 00009: val_accuracy did not improve from 0.30500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 2.0726 - accuracy: 0.2900 - val_loss: 1.9651 - val_accuracy: 0.3000\n",
      "Epoch 10/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 2.0507 - accuracy: 0.2946\n",
      "Epoch 00010: val_accuracy did not improve from 0.30500\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 2.0429 - accuracy: 0.2967 - val_loss: 1.9891 - val_accuracy: 0.2700\n",
      "Epoch 11/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.9269 - accuracy: 0.2969\n",
      "Epoch 00011: val_accuracy improved from 0.30500 to 0.35000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.9481 - accuracy: 0.3033 - val_loss: 1.8767 - val_accuracy: 0.3500\n",
      "Epoch 12/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.9075 - accuracy: 0.3419\n",
      "Epoch 00012: val_accuracy did not improve from 0.35000\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.9045 - accuracy: 0.3467 - val_loss: 1.8170 - val_accuracy: 0.3450\n",
      "Epoch 13/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.8359 - accuracy: 0.3361\n",
      "Epoch 00013: val_accuracy improved from 0.35000 to 0.36000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.8317 - accuracy: 0.3350 - val_loss: 1.8082 - val_accuracy: 0.3600\n",
      "Epoch 14/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.8387 - accuracy: 0.3301\n",
      "Epoch 00014: val_accuracy did not improve from 0.36000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.8510 - accuracy: 0.3250 - val_loss: 1.7806 - val_accuracy: 0.3450\n",
      "Epoch 15/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.7272 - accuracy: 0.3542\n",
      "Epoch 00015: val_accuracy did not improve from 0.36000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.7314 - accuracy: 0.3533 - val_loss: 1.7254 - val_accuracy: 0.3350\n",
      "Epoch 16/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.6253 - accuracy: 0.3589\n",
      "Epoch 00016: val_accuracy improved from 0.36000 to 0.37500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.6388 - accuracy: 0.3583 - val_loss: 1.6736 - val_accuracy: 0.3750\n",
      "Epoch 17/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.6256 - accuracy: 0.3946\n",
      "Epoch 00017: val_accuracy improved from 0.37500 to 0.40500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.6335 - accuracy: 0.3917 - val_loss: 1.6510 - val_accuracy: 0.4050\n",
      "Epoch 18/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.6144 - accuracy: 0.3682\n",
      "Epoch 00018: val_accuracy did not improve from 0.40500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.6175 - accuracy: 0.3667 - val_loss: 1.6520 - val_accuracy: 0.3850\n",
      "Epoch 19/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 1.6646 - accuracy: 0.3683\n",
      "Epoch 00019: val_accuracy did not improve from 0.40500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.6646 - accuracy: 0.3683 - val_loss: 1.6259 - val_accuracy: 0.3800\n",
      "Epoch 20/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.6446 - accuracy: 0.3767\n",
      "Epoch 00020: val_accuracy improved from 0.40500 to 0.42500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 1s 14ms/step - loss: 1.6418 - accuracy: 0.3800 - val_loss: 1.6385 - val_accuracy: 0.4250\n",
      "Epoch 21/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.5852 - accuracy: 0.3897\n",
      "Epoch 00021: val_accuracy improved from 0.42500 to 0.43500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.5767 - accuracy: 0.4017 - val_loss: 1.5721 - val_accuracy: 0.4350\n",
      "Epoch 22/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 1.5779 - accuracy: 0.4233\n",
      "Epoch 00022: val_accuracy did not improve from 0.43500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.5779 - accuracy: 0.4233 - val_loss: 1.6090 - val_accuracy: 0.3950\n",
      "Epoch 23/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.5963 - accuracy: 0.3750\n",
      "Epoch 00023: val_accuracy improved from 0.43500 to 0.44500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.5905 - accuracy: 0.3850 - val_loss: 1.5892 - val_accuracy: 0.4450\n",
      "Epoch 24/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.5294 - accuracy: 0.4277\n",
      "Epoch 00024: val_accuracy did not improve from 0.44500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.5427 - accuracy: 0.4167 - val_loss: 1.5796 - val_accuracy: 0.4150\n",
      "Epoch 25/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.5121 - accuracy: 0.4265\n",
      "Epoch 00025: val_accuracy did not improve from 0.44500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.5134 - accuracy: 0.4317 - val_loss: 1.5457 - val_accuracy: 0.4250\n",
      "Epoch 26/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.4918 - accuracy: 0.4286\n",
      "Epoch 00026: val_accuracy improved from 0.44500 to 0.46000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.4826 - accuracy: 0.4317 - val_loss: 1.5003 - val_accuracy: 0.4600\n",
      "Epoch 27/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.4442 - accuracy: 0.4706\n",
      "Epoch 00027: val_accuracy did not improve from 0.46000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.4563 - accuracy: 0.4667 - val_loss: 1.4900 - val_accuracy: 0.4450\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 28/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 1.4810 - accuracy: 0.4545\n",
      "Epoch 00028: val_accuracy improved from 0.46000 to 0.47000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.4721 - accuracy: 0.4583 - val_loss: 1.4937 - val_accuracy: 0.4700\n",
      "Epoch 29/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.4526 - accuracy: 0.4611\n",
      "Epoch 00029: val_accuracy improved from 0.47000 to 0.48500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 1.4546 - accuracy: 0.4600 - val_loss: 1.5133 - val_accuracy: 0.4850\n",
      "Epoch 30/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.4555 - accuracy: 0.4750\n",
      "Epoch 00030: val_accuracy did not improve from 0.48500\n",
      "38/38 [==============================] - 1s 13ms/step - loss: 1.4636 - accuracy: 0.4717 - val_loss: 1.5071 - val_accuracy: 0.4400\n",
      "Epoch 31/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.3523 - accuracy: 0.4661\n",
      "Epoch 00031: val_accuracy did not improve from 0.48500\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 1.3662 - accuracy: 0.4600 - val_loss: 1.4698 - val_accuracy: 0.4600\n",
      "Epoch 32/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.3814 - accuracy: 0.4792\n",
      "Epoch 00032: val_accuracy did not improve from 0.48500\n",
      "38/38 [==============================] - 1s 17ms/step - loss: 1.3857 - accuracy: 0.4800 - val_loss: 1.4983 - val_accuracy: 0.4650\n",
      "Epoch 33/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.3832 - accuracy: 0.5118\n",
      "Epoch 00033: val_accuracy did not improve from 0.48500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.3852 - accuracy: 0.5083 - val_loss: 1.4807 - val_accuracy: 0.4550\n",
      "Epoch 34/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.3313 - accuracy: 0.5107\n",
      "Epoch 00034: val_accuracy improved from 0.48500 to 0.49000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.3128 - accuracy: 0.5200 - val_loss: 1.3765 - val_accuracy: 0.4900\n",
      "Epoch 35/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 1.3099 - accuracy: 0.4943\n",
      "Epoch 00035: val_accuracy improved from 0.49000 to 0.51000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.3145 - accuracy: 0.4883 - val_loss: 1.3363 - val_accuracy: 0.5100\n",
      "Epoch 36/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.2847 - accuracy: 0.5404\n",
      "Epoch 00036: val_accuracy improved from 0.51000 to 0.51500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.2899 - accuracy: 0.5383 - val_loss: 1.3581 - val_accuracy: 0.5150\n",
      "Epoch 37/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 1.2920 - accuracy: 0.5076\n",
      "Epoch 00037: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.2867 - accuracy: 0.5133 - val_loss: 1.4343 - val_accuracy: 0.5100\n",
      "Epoch 38/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.3255 - accuracy: 0.4929\n",
      "Epoch 00038: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 1s 20ms/step - loss: 1.2958 - accuracy: 0.5033 - val_loss: 1.3388 - val_accuracy: 0.5050\n",
      "Epoch 39/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.2874 - accuracy: 0.5464\n",
      "Epoch 00039: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 1s 19ms/step - loss: 1.2935 - accuracy: 0.5450 - val_loss: 1.3443 - val_accuracy: 0.4850\n",
      "Epoch 40/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.2418 - accuracy: 0.5589\n",
      "Epoch 00040: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.2418 - accuracy: 0.5583 - val_loss: 1.3630 - val_accuracy: 0.5000\n",
      "Epoch 41/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.2576 - accuracy: 0.5590\n",
      "Epoch 00041: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.2501 - accuracy: 0.5567 - val_loss: 1.3325 - val_accuracy: 0.4900\n",
      "Epoch 42/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.1954 - accuracy: 0.5801 ETA: 0s - loss: 1.1715 - accu\n",
      "Epoch 00042: val_accuracy did not improve from 0.51500\n",
      "38/38 [==============================] - 1s 18ms/step - loss: 1.1975 - accuracy: 0.5767 - val_loss: 1.3176 - val_accuracy: 0.5100\n",
      "Epoch 43/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.2352 - accuracy: 0.5226\n",
      "Epoch 00043: val_accuracy improved from 0.51500 to 0.52000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.2278 - accuracy: 0.5267 - val_loss: 1.2884 - val_accuracy: 0.5200\n",
      "Epoch 44/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.1968 - accuracy: 0.5176\n",
      "Epoch 00044: val_accuracy improved from 0.52000 to 0.53500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.2018 - accuracy: 0.5233 - val_loss: 1.3055 - val_accuracy: 0.5350\n",
      "Epoch 45/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.1743 - accuracy: 0.5486\n",
      "Epoch 00045: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.1588 - accuracy: 0.5550 - val_loss: 1.3145 - val_accuracy: 0.5000\n",
      "Epoch 46/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 1.1740 - accuracy: 0.5767\n",
      "Epoch 00046: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 1.1740 - accuracy: 0.5767 - val_loss: 1.2733 - val_accuracy: 0.4950\n",
      "Epoch 47/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 1.1514 - accuracy: 0.5733\n",
      "Epoch 00047: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 1.1514 - accuracy: 0.5733 - val_loss: 1.2674 - val_accuracy: 0.5250\n",
      "Epoch 48/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.1371 - accuracy: 0.5821\n",
      "Epoch 00048: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 1.1290 - accuracy: 0.5850 - val_loss: 1.2872 - val_accuracy: 0.5200\n",
      "Epoch 49/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 1.1091 - accuracy: 0.5963\n",
      "Epoch 00049: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 1s 14ms/step - loss: 1.1041 - accuracy: 0.5967 - val_loss: 1.2728 - val_accuracy: 0.5350\n",
      "Epoch 50/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 1.0972 - accuracy: 0.5816\n",
      "Epoch 00050: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 1s 14ms/step - loss: 1.0947 - accuracy: 0.5817 - val_loss: 1.2248 - val_accuracy: 0.5100\n",
      "Epoch 51/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.0800 - accuracy: 0.6029\n",
      "Epoch 00051: val_accuracy did not improve from 0.53500\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.1014 - accuracy: 0.6000 - val_loss: 1.2083 - val_accuracy: 0.5300\n",
      "Epoch 52/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 1.1337 - accuracy: 0.5947\n",
      "Epoch 00052: val_accuracy improved from 0.53500 to 0.54500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.0978 - accuracy: 0.6033 - val_loss: 1.2960 - val_accuracy: 0.5450\n",
      "Epoch 53/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.1167 - accuracy: 0.6048\n",
      "Epoch 00053: val_accuracy improved from 0.54500 to 0.56000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 1.1128 - accuracy: 0.6000 - val_loss: 1.1753 - val_accuracy: 0.5600\n",
      "Epoch 54/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.0583 - accuracy: 0.6121\n",
      "Epoch 00054: val_accuracy improved from 0.56000 to 0.56500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.0434 - accuracy: 0.6233 - val_loss: 1.2215 - val_accuracy: 0.5650\n",
      "Epoch 55/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.0448 - accuracy: 0.6048\n",
      "Epoch 00055: val_accuracy improved from 0.56500 to 0.57500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 1.0635 - accuracy: 0.5983 - val_loss: 1.1862 - val_accuracy: 0.5750\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 56/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 1.0278 - accuracy: 0.6232\n",
      "Epoch 00056: val_accuracy did not improve from 0.57500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.0386 - accuracy: 0.6183 - val_loss: 1.1993 - val_accuracy: 0.5700\n",
      "Epoch 57/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 1.0202 - accuracy: 0.6140\n",
      "Epoch 00057: val_accuracy improved from 0.57500 to 0.60000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.0148 - accuracy: 0.6167 - val_loss: 1.1859 - val_accuracy: 0.6000\n",
      "Epoch 58/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.9985 - accuracy: 0.6500\n",
      "Epoch 00058: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 1.0113 - accuracy: 0.6417 - val_loss: 1.1821 - val_accuracy: 0.5700\n",
      "Epoch 59/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.9890 - accuracy: 0.6458\n",
      "Epoch 00059: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.9920 - accuracy: 0.6433 - val_loss: 1.1930 - val_accuracy: 0.5600\n",
      "Epoch 60/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.9789 - accuracy: 0.6402\n",
      "Epoch 00060: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.9799 - accuracy: 0.6367 - val_loss: 1.2379 - val_accuracy: 0.5500\n",
      "Epoch 61/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 1.0214 - accuracy: 0.6191\n",
      "Epoch 00061: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.9899 - accuracy: 0.6283 - val_loss: 1.1880 - val_accuracy: 0.5650\n",
      "Epoch 62/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.9478 - accuracy: 0.6420\n",
      "Epoch 00062: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.9385 - accuracy: 0.6450 - val_loss: 1.2178 - val_accuracy: 0.5750\n",
      "Epoch 63/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 0.9597 - accuracy: 0.6334 ETA: 0s - loss: 0.9578 - accuracy: 0.\n",
      "Epoch 00063: val_accuracy did not improve from 0.60000\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 0.9529 - accuracy: 0.6367 - val_loss: 1.2034 - val_accuracy: 0.5550\n",
      "Epoch 64/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.9474 - accuracy: 0.6389\n",
      "Epoch 00064: val_accuracy improved from 0.60000 to 0.62000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 1s 17ms/step - loss: 0.9507 - accuracy: 0.6350 - val_loss: 1.1667 - val_accuracy: 0.6200\n",
      "Epoch 65/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.9677 - accuracy: 0.6528\n",
      "Epoch 00065: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 0.9707 - accuracy: 0.6500 - val_loss: 1.1371 - val_accuracy: 0.5800\n",
      "Epoch 66/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 0.8905 - accuracy: 0.6719\n",
      "Epoch 00066: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 0.9011 - accuracy: 0.6733 - val_loss: 1.1605 - val_accuracy: 0.5800\n",
      "Epoch 67/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.8991 - accuracy: 0.6679\n",
      "Epoch 00067: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.9076 - accuracy: 0.6633 - val_loss: 1.1249 - val_accuracy: 0.6000\n",
      "Epoch 68/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 0.8560 - accuracy: 0.6824\n",
      "Epoch 00068: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.8544 - accuracy: 0.6833 - val_loss: 1.1365 - val_accuracy: 0.5900\n",
      "Epoch 69/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 0.8451 - accuracy: 0.6895\n",
      "Epoch 00069: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.8534 - accuracy: 0.6900 - val_loss: 1.1558 - val_accuracy: 0.6000\n",
      "Epoch 70/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.8552 - accuracy: 0.6673\n",
      "Epoch 00070: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.8621 - accuracy: 0.6667 - val_loss: 1.1104 - val_accuracy: 0.5850\n",
      "Epoch 71/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.8368 - accuracy: 0.6970\n",
      "Epoch 00071: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.8249 - accuracy: 0.6983 - val_loss: 1.1197 - val_accuracy: 0.6150\n",
      "Epoch 72/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 0.8699 - accuracy: 0.6836\n",
      "Epoch 00072: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.8656 - accuracy: 0.6900 - val_loss: 1.1089 - val_accuracy: 0.6200\n",
      "Epoch 73/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 0.8520 - accuracy: 0.6892\n",
      "Epoch 00073: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.8512 - accuracy: 0.6900 - val_loss: 1.0975 - val_accuracy: 0.6200\n",
      "Epoch 74/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 0.8353 - accuracy: 0.6883\n",
      "Epoch 00074: val_accuracy did not improve from 0.62000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.8353 - accuracy: 0.6883 - val_loss: 1.1435 - val_accuracy: 0.5550\n",
      "Epoch 75/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.8308 - accuracy: 0.6970\n",
      "Epoch 00075: val_accuracy improved from 0.62000 to 0.63000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.8439 - accuracy: 0.6833 - val_loss: 1.1189 - val_accuracy: 0.6300\n",
      "Epoch 76/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 0.8610 - accuracy: 0.6867\n",
      "Epoch 00076: val_accuracy improved from 0.63000 to 0.63500, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.8610 - accuracy: 0.6867 - val_loss: 1.1423 - val_accuracy: 0.6350\n",
      "Epoch 77/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 0.7932 - accuracy: 0.6993\n",
      "Epoch 00077: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.7937 - accuracy: 0.7000 - val_loss: 1.2111 - val_accuracy: 0.5850\n",
      "Epoch 78/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 0.8037 - accuracy: 0.7051\n",
      "Epoch 00078: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.7819 - accuracy: 0.7167 - val_loss: 1.1078 - val_accuracy: 0.6300\n",
      "Epoch 79/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.7727 - accuracy: 0.7196\n",
      "Epoch 00079: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.7647 - accuracy: 0.7200 - val_loss: 1.1107 - val_accuracy: 0.5900\n",
      "Epoch 80/100\n",
      "38/38 [==============================] - ETA: 0s - loss: 0.7315 - accuracy: 0.7317\n",
      "Epoch 00080: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.7315 - accuracy: 0.7317 - val_loss: 1.1329 - val_accuracy: 0.5950\n",
      "Epoch 81/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.7371 - accuracy: 0.7224\n",
      "Epoch 00081: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 0.7303 - accuracy: 0.7250 - val_loss: 1.0859 - val_accuracy: 0.6050\n",
      "Epoch 82/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.7911 - accuracy: 0.6967\n",
      "Epoch 00082: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 1s 17ms/step - loss: 0.7871 - accuracy: 0.6983 - val_loss: 1.1484 - val_accuracy: 0.6100\n",
      "Epoch 83/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.6927 - accuracy: 0.7661\n",
      "Epoch 00083: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 1s 21ms/step - loss: 0.7108 - accuracy: 0.7600 - val_loss: 1.1643 - val_accuracy: 0.6150\n",
      "Epoch 84/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "36/38 [===========================>..] - ETA: 0s - loss: 0.6988 - accuracy: 0.7344\n",
      "Epoch 00084: val_accuracy did not improve from 0.63500\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.7097 - accuracy: 0.7283 - val_loss: 1.1450 - val_accuracy: 0.6000\n",
      "Epoch 85/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.7139 - accuracy: 0.7292\n",
      "Epoch 00085: val_accuracy improved from 0.63500 to 0.65000, saving model to models/simple_dnn/model.h5\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.7056 - accuracy: 0.7367 - val_loss: 1.0560 - val_accuracy: 0.6500\n",
      "Epoch 86/100\n",
      "37/38 [============================>.] - ETA: 0s - loss: 0.7031 - accuracy: 0.7348\n",
      "Epoch 00086: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.7041 - accuracy: 0.7333 - val_loss: 1.1320 - val_accuracy: 0.6200\n",
      "Epoch 87/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.6615 - accuracy: 0.7595\n",
      "Epoch 00087: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.6674 - accuracy: 0.7533 - val_loss: 1.1412 - val_accuracy: 0.6200\n",
      "Epoch 88/100\n",
      "32/38 [========================>.....] - ETA: 0s - loss: 0.7196 - accuracy: 0.7363\n",
      "Epoch 00088: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.7004 - accuracy: 0.7483 - val_loss: 1.1172 - val_accuracy: 0.5950\n",
      "Epoch 89/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.6763 - accuracy: 0.7552\n",
      "Epoch 00089: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 13ms/step - loss: 0.6787 - accuracy: 0.7483 - val_loss: 1.0932 - val_accuracy: 0.6000\n",
      "Epoch 90/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.6809 - accuracy: 0.7464\n",
      "Epoch 00090: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 0.6889 - accuracy: 0.7450 - val_loss: 1.1438 - val_accuracy: 0.6200\n",
      "Epoch 91/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.7005 - accuracy: 0.7261\n",
      "Epoch 00091: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 1s 16ms/step - loss: 0.7059 - accuracy: 0.7200 - val_loss: 1.1267 - val_accuracy: 0.6250\n",
      "Epoch 92/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.6599 - accuracy: 0.7569\n",
      "Epoch 00092: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.6471 - accuracy: 0.7633 - val_loss: 1.2089 - val_accuracy: 0.6050\n",
      "Epoch 93/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.6842 - accuracy: 0.7390\n",
      "Epoch 00093: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.6694 - accuracy: 0.7417 - val_loss: 1.1494 - val_accuracy: 0.6200\n",
      "Epoch 94/100\n",
      "34/38 [=========================>....] - ETA: 0s - loss: 0.6275 - accuracy: 0.7757\n",
      "Epoch 00094: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.6236 - accuracy: 0.7767 - val_loss: 1.1624 - val_accuracy: 0.6050\n",
      "Epoch 95/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.6241 - accuracy: 0.7679\n",
      "Epoch 00095: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 1s 14ms/step - loss: 0.6168 - accuracy: 0.7700 - val_loss: 1.1092 - val_accuracy: 0.6400\n",
      "Epoch 96/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.6133 - accuracy: 0.7679\n",
      "Epoch 00096: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 12ms/step - loss: 0.6188 - accuracy: 0.7683 - val_loss: 1.1570 - val_accuracy: 0.6150\n",
      "Epoch 97/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.6201 - accuracy: 0.7604\n",
      "Epoch 00097: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 1s 15ms/step - loss: 0.6204 - accuracy: 0.7617 - val_loss: 1.1107 - val_accuracy: 0.6350\n",
      "Epoch 98/100\n",
      "33/38 [=========================>....] - ETA: 0s - loss: 0.5584 - accuracy: 0.7898\n",
      "Epoch 00098: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.5847 - accuracy: 0.7850 - val_loss: 1.1346 - val_accuracy: 0.6450\n",
      "Epoch 99/100\n",
      "36/38 [===========================>..] - ETA: 0s - loss: 0.5770 - accuracy: 0.7951\n",
      "Epoch 00099: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 1s 17ms/step - loss: 0.5794 - accuracy: 0.7933 - val_loss: 1.2074 - val_accuracy: 0.6250\n",
      "Epoch 100/100\n",
      "35/38 [==========================>...] - ETA: 0s - loss: 0.5941 - accuracy: 0.7607\n",
      "Epoch 00100: val_accuracy did not improve from 0.65000\n",
      "38/38 [==============================] - 0s 11ms/step - loss: 0.5985 - accuracy: 0.7667 - val_loss: 1.1758 - val_accuracy: 0.6150\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train,\n",
    "    y_train,\n",
    "    epochs=100,\n",
    "    batch_size=16,\n",
    "    validation_data=(X_valid, y_valid),\n",
    "    callbacks=[\n",
    "        ModelCheckpoint(MODEL_PATH, monitor='val_accuracy', mode='max', verbose=1, save_best_only=True),\n",
    "#         EarlyStopping(monitor='val_loss', mode='max', verbose=1, patience=5, min_delta=0.01)\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 576x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_loss_accuracy(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7/7 [==============================] - 0s 10ms/step - loss: 1.5918 - accuracy: 0.5350\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[1.591825008392334, 0.5350000262260437]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}