train_tomoscope.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train the ML model\n",
    "from models import Tomoscope\n",
    "from utils import plot_loss\n",
    "from utils import fast_tensor_load_encdec\n",
    "\n",
    "import time\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "import yaml\n",
    "import os\n",
    "import numpy as np\n",
    "from datetime import datetime\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using timestamp:  tomoscope_test\n"
     ]
    }
   ],
   "source": [
    "var_names = ['phEr', 'enEr', 'bl',\n",
    "             'inten', 'Vrf', 'mu', 'VrfSPS']\n",
    "# Initialize parameters\n",
    "data_dir = './tomo_data/datasets_tomoscope_TF_24-03-23/'\n",
    "\n",
    "timestamp = datetime.now().strftime(\"%Y_%m_%d_%H-%M-%S\")\n",
    "timestamp = 'tomoscope_test'\n",
    "print('Using timestamp: ', timestamp)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data specific\n",
    "IMG_OUTPUT_SIZE = 128\n",
    "\n",
    "# Train specific\n",
    "# train_cfg = {\n",
    "#     'epochs': 10, 'output_turns': 10,\n",
    "#     'cropping': [0, 0],\n",
    "#     'enc_dense_layers': [1024, 256],\n",
    "#     'dec_dense_layers': [1024],\n",
    "#     'enc_filters': [4, 8, 16],\n",
    "#     'dec_filters': [8, 16, 10],\n",
    "#     'enc_kernel_size': 3,\n",
    "#     'dec_kernel_size': 3,\n",
    "#     'enc_strides': [2, 2],\n",
    "#     'dec_strides': [2, 2],\n",
    "#     'enc_activation': 'relu',\n",
    "#     'dec_activation': 'relu',\n",
    "#     'enc_pooling': None, 'dec_pooling': None,\n",
    "#     'enc_pooling_size': [0, 0], 'dec_pooling_size': [0, 0],\n",
    "#     'enc_pooling_strides': [1, 1], 'dec_pooling_strides': [1, 1],\n",
    "#     'enc_pooling_padding': 'valid', 'dec_pooling_padding': 'valid',\n",
    "#     'enc_dropout': 0.0, 'dec_dropout': 0.0,\n",
    "#     'metrics': [], 'use_bias': False, 'batchnorm': False,\n",
    "#     'learning_rate': 1e-3,\n",
    "#     'dataset%': 0.1,\n",
    "#     'normalization': 'minmax', 'img_normalize': 'off',\n",
    "#     'ps_normalize': 'off',\n",
    "#     'batch_size': 32\n",
    "# }\n",
    "\n",
    "\n",
    "# Train specific\n",
    "train_cfg = {\n",
    "    'epochs': 10, 'output_turns': 1,\n",
    "    'cropping': [0, 0],\n",
    "    'enc_filters': [16, 32, 64, 128],\n",
    "    'dec_filters': [64, 32, 16],\n",
    "    'enc_kernel_size': 4,\n",
    "    'dec_kernel_size': 4,\n",
    "    'enc_strides': [2, 2],\n",
    "    'dec_strides': [2, 2],\n",
    "    'enc_activation': 'relu',\n",
    "    'dec_activation': 'relu',\n",
    "    'enc_conv_padding': 'same', 'dec_conv_padding': 'same',\n",
    "    'enc_dropout': 0.0, 'dec_dropout': 0.0,\n",
    "    'metrics': [], 'use_bias': False, 'batchnorm': False,\n",
    "    'learning_rate': 1e-3,\n",
    "    'dataset%': 0.1,\n",
    "    'normalization': 'minmax', 'img_normalize': 'off',\n",
    "    'ps_normalize': 'off',\n",
    "    'batch_size': 32\n",
    "}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading ./tomo_data/datasets_tomoscope_TF_24-03-23/ML_data/tomoscope-training-00.npz\n",
      "Loading ./tomo_data/datasets_tomoscope_TF_24-03-23/ML_data/tomoscope-training-01.npz\n",
      "Loading ./tomo_data/datasets_tomoscope_TF_24-03-23/ML_data/tomoscope-training-02.npz\n",
      "Loading ./tomo_data/datasets_tomoscope_TF_24-03-23/ML_data/tomoscope-training-03.npz\n",
      "Number of Training files:  1589\n",
      "Loading ./tomo_data/datasets_tomoscope_TF_24-03-23/ML_data/tomoscope-validation-00.npz\n",
      "Number of Validation files:  140\n"
     ]
    }
   ],
   "source": [
    "# Initialize directories\n",
    "trial_dir = os.path.join('./trials/', timestamp)\n",
    "weights_dir = os.path.join(trial_dir, 'weights')\n",
    "plots_dir = os.path.join(trial_dir, 'plots')\n",
    "\n",
    "# Initialize train/ test / validation paths\n",
    "ML_dir = os.path.join(data_dir, 'ML_data')\n",
    "assert os.path.exists(ML_dir)\n",
    "\n",
    "TRAINING_PATH = os.path.join(ML_dir, 'TRAINING')\n",
    "VALIDATION_PATH = os.path.join(ML_dir, 'VALIDATION')\n",
    "\n",
    "# create the directory to store the results\n",
    "os.makedirs(trial_dir, exist_ok=True)\n",
    "os.makedirs(weights_dir, exist_ok=True)\n",
    "os.makedirs(plots_dir, exist_ok=True)\n",
    "\n",
    "\n",
    "assert train_cfg['normalization'] == 'minmax'\n",
    "assert train_cfg['img_normalize'] == 'off'\n",
    "assert train_cfg['ps_normalize'] == 'off'\n",
    "\n",
    "TRAINING_PATH = os.path.join(ML_dir, 'tomoscope-training-??.npz')\n",
    "VALIDATION_PATH = os.path.join(ML_dir, 'tomoscope-validation-??.npz')\n",
    "\n",
    "x_train, turn_train, latent_train, y_train = fast_tensor_load_encdec(\n",
    "    TRAINING_PATH, train_cfg['dataset%'])\n",
    "print('Number of Training files: ', len(y_train))\n",
    "\n",
    "x_valid, turn_valid, latent_valid, y_valid = fast_tensor_load_encdec(\n",
    "    VALIDATION_PATH, train_cfg['dataset%'])\n",
    "print('Number of Validation files: ', len(y_valid))\n",
    "\n",
    "y_train = y_train[:, :, :, :train_cfg['output_turns']]\n",
    "y_valid = y_valid[:, :, :, :train_cfg['output_turns']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training data shape:  (1589, 128, 128, 1) (1589, 128, 128, 1)\n",
      "turns data shape:  (1589, 10)\n",
      "latent data shape:  (1589, 7)\n"
     ]
    }
   ],
   "source": [
    "print('Training data shape: ', x_train.shape, y_train.shape)\n",
    "print('turns data shape: ', turn_train.shape)\n",
    "print('latent data shape: ', latent_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "# plot some of the outputs\n",
    "\n",
    "nrows = 3\n",
    "# Get nrows * nrows random images\n",
    "sample = np.random.choice(np.arange(len(x_train)),\n",
    "                          size=nrows * nrows, replace=False)\n",
    "\n",
    "samples_X = tf.gather(x_train, sample)\n",
    "samples_y = tf.gather(y_train, sample)\n",
    "\n",
    "# Create 3x3 grid of figures\n",
    "fig, axes = plt.subplots(ncols=nrows, nrows=nrows, figsize=(12, 12))\n",
    "axes = np.ravel(axes)\n",
    "for i in range(len(axes)):\n",
    "    ax = axes[i]\n",
    "    ax.set_xticks([])\n",
    "    ax.set_yticks([])\n",
    "    # show the image\n",
    "    ax.imshow(samples_X[i], cmap='jet')\n",
    "    # Set the label\n",
    "    # title = ','.join([f'{num:.1f}' for num in samples_y[i]])\n",
    "    # ax.set_title(f'{title}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "---- Initializing model ----\n",
      "\n",
      "Model: \"model_7\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      " input (InputLayer)             [(None, 128, 128, 1  0           []                               \n",
      "                                )]                                                                \n",
      "                                                                                                  \n",
      " crop (Cropping2D)              (None, 128, 128, 1)  0           ['input[0][0]']                  \n",
      "                                                                                                  \n",
      " encoder_cnn_1 (Sequential)     (None, 64, 64, 16)   256         ['crop[0][0]']                   \n",
      "                                                                                                  \n",
      " encoder_cnn_2 (Sequential)     (None, 32, 32, 32)   8192        ['encoder_cnn_1[0][0]']          \n",
      "                                                                                                  \n",
      " encoder_cnn_3 (Sequential)     (None, 16, 16, 64)   32768       ['encoder_cnn_2[0][0]']          \n",
      "                                                                                                  \n",
      " encoder_cnn_4 (Sequential)     (None, 8, 8, 128)    131072      ['encoder_cnn_3[0][0]']          \n",
      "                                                                                                  \n",
      " decoder_cnn_1 (Sequential)     (None, 16, 16, 64)   131328      ['encoder_cnn_4[0][0]']          \n",
      "                                                                                                  \n",
      " concat_1 (Concatenate)         (None, 16, 16, 128)  0           ['decoder_cnn_1[0][0]',          \n",
      "                                                                  'encoder_cnn_3[0][0]']          \n",
      "                                                                                                  \n",
      " decoder_cnn_2 (Sequential)     (None, 32, 32, 32)   65664       ['concat_1[0][0]']               \n",
      "                                                                                                  \n",
      " concat_2 (Concatenate)         (None, 32, 32, 64)   0           ['decoder_cnn_2[0][0]',          \n",
      "                                                                  'encoder_cnn_2[0][0]']          \n",
      "                                                                                                  \n",
      " decoder_cnn_3 (Sequential)     (None, 64, 64, 16)   16448       ['concat_2[0][0]']               \n",
      "                                                                                                  \n",
      " concat_3 (Concatenate)         (None, 64, 64, 32)   0           ['decoder_cnn_3[0][0]',          \n",
      "                                                                  'encoder_cnn_1[0][0]']          \n",
      "                                                                                                  \n",
      " decoder_cnn_4 (Conv2DTranspose  (None, 128, 128, 1)  513        ['concat_3[0][0]']               \n",
      " )                                                                                                \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 386,241\n",
      "Trainable params: 386,017\n",
      "Non-trainable params: 224\n",
      "__________________________________________________________________________________________________\n",
      "None\n",
      "\n",
      "---- Model has been initialized, elapsed: 0.16292166709899902 ----\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from models import Tomoscope\n",
    "# Model instantiation\n",
    "start_t = time.time()\n",
    "input_shape = x_train.shape[1:]\n",
    "models = {}\n",
    "\n",
    "print(f'\\n---- Initializing model ----\\n')\n",
    "\n",
    "tomoscope = Tomoscope(input_shape=input_shape, **train_cfg)\n",
    "print(tomoscope.model.summary())\n",
    "print(\n",
    "    f'\\n---- Model has been initialized, elapsed: {time.time() - start_t} ----\\n')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.utils.plot_model(tomoscope.model, show_shapes=True, dpi=64)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "print(f'\\n---- Training the model ----\\n')\n",
    "\n",
    "# callbacks, save the best model, and early stop if no improvement in val_loss\n",
    "stop_early = keras.callbacks.EarlyStopping(monitor='val_loss',\n",
    "                                            patience=10, restore_best_weights=True)\n",
    "save_best = keras.callbacks.ModelCheckpoint(filepath=os.path.join(weights_dir, f'tomoscope.h5'),\n",
    "                                            monitor='val_loss', save_best_only=True)\n",
    "\n",
    "start_time = time.time()\n",
    "\n",
    "history = tomoscope.model.fit(\n",
    "    x=x_train, y=y_train, \n",
    "    epochs=train_cfg['epochs'],\n",
    "    validation_data=(x_valid, y_valid),\n",
    "    callbacks=[save_best], \n",
    "    batch_size=train_cfg['batch_size'],\n",
    "    verbose=0)\n",
    "\n",
    "total_time = time.time() - start_time\n",
    "print(\n",
    "    f'\\n---- Training complete, epochs: {len(history.history[\"loss\"])}, min loss {np.min(history.history[\"val_loss\"])}, total time {total_time} ----\\n')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor(0.8296597, shape=(), dtype=float32)\n",
      "0.008987858\n"
     ]
    }
   ],
   "source": [
    "print(custom_loss(y_train[0:1], tomoscope.model(x_train[0:1])))\n",
    "print(np.mean(tf.keras.metrics.mean_squared_error(y_train[0:1], tomoscope.model(x_train[0:1]))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "---- Plotting loss ----\n",
      "\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "# Plot training and validation loss\n",
    "print('\\n---- Plotting loss ----\\n')\n",
    "\n",
    "train_loss_l = np.array(history.history['loss'])\n",
    "valid_loss_l = np.array(history.history['val_loss'])\n",
    "\n",
    "plot_loss({'Training': train_loss_l, 'Validation': valid_loss_l},\n",
    "            title='Tomoscpe Train/Validation Loss',\n",
    "            figname=os.path.join(plots_dir, 'tomoscope_train_valid_loss.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save file with experiment configuration\n",
    "config_dict = {}\n",
    "config_dict['tomoscope'] = train_cfg.copy()\n",
    "config_dict['tomoscope'].update({\n",
    "    'min_train_loss': float(np.min(train_loss_l)),\n",
    "    'min_valid_loss': float(np.min(valid_loss_l)),\n",
    "    'total_train_time': total_time,\n",
    "    'used_gpus': 0\n",
    "})\n",
    "\n",
    "# save config_dict\n",
    "with open(os.path.join(trial_dir, 'tomoscope-summary.yml'), 'w') as configfile:\n",
    "    yaml.dump(config_dict, configfile, default_flow_style=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.9.7"
  },
  "vscode": {
   "interpreter": {
    "hash": "79ab8fd01a8cec42884b8b2a5d7fb4751c5402d97e9e61d151ed5c6a6352873c"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}