{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This example demonstrates the whole process from initial atomic structure to training, evaluation and prediction. It includes:\n",
    "\n",
    "\n",
    "1. Read input atomic structures (saved as extxyz files) and create descriptors and their derivatives.\n",
    "\n",
    "2. Read inputs and outputs into a Data object.\n",
    "\n",
    "3. Create tensorflow dataset for training.\n",
    "\n",
    "4. Train the potential and apply it for prediction.\n",
    "\n",
    "5. Save the trained model and then load it for retraining or prediction.\n",
    "\n",
    "\n",
    "The code has been tested on Tensorflow 2.5 and 2.6."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import atomdnn\n",
    "\n",
    "# 'float64' is used for reading data and train by default\n",
    "atomdnn.data_type = 'float64'\n",
    "\n",
    "# force and stress are evaluated by default, \n",
    "# if one only need to compute potential energy, then set compute_force to false\n",
    "atomdnn.compute_force = True\n",
    "\n",
    "# default value is for converting ev/A^3 to GPa\n",
    "# note that: the predicted positive stress means tension and negative stress means compression\n",
    "stress_unit_convert = 160.2176 \n",
    "\n",
    "import numpy as np\n",
    "import pickle\n",
    "import tensorflow as tf\n",
    "from atomdnn import data\n",
    "from atomdnn.data import Data\n",
    "from atomdnn.data import *\n",
    "from atomdnn.descriptor import *\n",
    "from atomdnn import network\n",
    "from atomdnn.network import Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create descriptors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Read input atomic structures (saved as extxyz files) and create descriptors and their derivatives**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total number of fingerprints = 14\n"
     ]
    }
   ],
   "source": [
    "descriptor = {'name': 'acsf', \n",
    "              'cutoff': 6.5,\n",
    "              'etaG2':[0.01,0.1,1,5,10], \n",
    "              'etaG4': [0.01], \n",
    "              'zeta': [0.08,1.0,10.0,100.0],\n",
    "              'lambda': [1.0, -1.0]}\n",
    "\n",
    "\n",
    "# define lammps excutable (serial or mpi) \n",
    "# LAMMPS has to be compiled with the added compute and dump_local subrutines (inside atomdnn/lammps)\n",
    "lmpexe = 'lmp_serial' \n",
    "#lmpexe = 'mpirun -np 2 lmp_mpi'  # can be mpi version\n",
    "\n",
    "\n",
    "elements = ['C']\n",
    "xyzfile_path='extxyz'\n",
    "xyzfile_name = 'example_extxyz.*' # a serials of files like example_extxyz.1, example_extxyz.2, ...example_extxyz.n\n",
    "descriptors_path = './descriptors'\n",
    "\n",
    "descriptor_filename = 'dump_fp.*' # a serials of dump_fp.* files will be created\n",
    "der_filename ='dump_der.*'\n",
    "\n",
    "print('total number of fingerprints = %i'%get_num_fingerprints(descriptor,elements))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start creating fingerprints and derivatives for 50 files ...\n",
      "  so far finished for 10 images ...\n",
      "  so far finished for 20 images ...\n",
      "  so far finished for 30 images ...\n",
      "  so far finished for 40 images ...\n",
      "  so far finished for 50 images ...\n",
      "Finish creating descriptors and derivatives for total 50 images.\n",
      "It took 10.55 seconds.\n"
     ]
    }
   ],
   "source": [
    "# this will create a serials of files for descriptors and their derivatives inside descriptors_path\n",
    "create_descriptors(lmpexe,\n",
    "                   elements,\n",
    "                   xyzfile_path, \n",
    "                   xyzfile_name, \n",
    "                   descriptors_path, \n",
    "                   descriptor, \n",
    "                   descriptor_filename, \n",
    "                   der_filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read inputs&outputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Read inputs and outputs into a Data object** "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a Data object\n",
    "mydata = Data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start reading fingerprints from 'dump_fp.*' for total 50 files ...\n",
      "  so far read 50 images ...\n",
      "  Finish reading fingerprints from total 50 images.\n",
      "\n",
      "\n",
      "Start reading derivatives from 'dump_der.*' for total 50 files ...\n",
      "  This may take a while for large data set ...\n",
      "  so far read 50 images ...\n",
      "  Finish reading dGdr derivatives from total 50 images.\n",
      "\n",
      "  It took 1.02 seconds to read the derivatives data.\n",
      "\n",
      "---------- input dataset information ----------\n",
      "total images = 50\n",
      "max number of atoms = 4\n",
      "number of fingerprints = 14\n",
      "number of atom types = 1\n",
      "max number of derivative pairs = 200\n",
      "------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "# read inputs: descriptors and their derivatives\n",
    "mydata.read_inputdata(descriptors_path, descriptor_filename, der_filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading outputs from extxyz files ...\n",
      "  so far read 50 images ...\n",
      "  Finish reading outputs from total 50 images.\n",
      "\n",
      "\n",
      "---------- output dataset information ------------\n",
      "total images = 50\n",
      "max number of atoms = 4\n",
      "read_force = True\n",
      "read_stress = True\n",
      "---------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "# read outputs: potential energy, force and stress from extxyz files\n",
    "mydata.read_outputdata(xyzfile_path, xyzfile_name)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create TFdataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Create tensorflow dataset for training**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Conversion may take a while for large datasets...\n",
      "It took 0.0046 second.\n"
     ]
    }
   ],
   "source": [
    "# convert data to tensors\n",
    "mydata.convert_data_to_tensor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create tensorflow dataset\n",
    "tf_dataset = tf.data.Dataset.from_tensor_slices((mydata.input_dict,mydata.output_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_path = './example_tfdataset'\n",
    "\n",
    "# save the dataset\n",
    "tf.data.experimental.save(tf_dataset, dataset_path)\n",
    "\n",
    "# save the element_spec to disk for future loading, this is only needed for tensorflow lower than 2.6\n",
    "with open(dataset_path + '/element_spec', 'wb') as out_: \n",
    "    pickle.dump(tf_dataset.element_spec, out_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note: The above three steps just need to be done once for one data set, the training only uses the saved tensorflow dataset.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Load the dataset and train the model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load tensorflow dataset, for Tensorflow version lower than 2.6, need to specify element_spec.\n",
    "\n",
    "with open(dataset_path + '/element_spec', 'rb') as in_:\n",
    "    element_spec = pickle.load(in_)\n",
    "\n",
    "dataset = tf.data.experimental.load(dataset_path,element_spec=element_spec)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Traning data: 35 images\n",
      "Validation data: 10 images\n",
      "Test data: 5 images\n"
     ]
    }
   ],
   "source": [
    "# split the data to training, validation and testing sets\n",
    "\n",
    "train_dataset, val_dataset, test_dataset = split_dataset(dataset,0.7,0.2,0.1,shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the network\n",
    "# See section 'Training' for detailed description on Network object.\n",
    "\n",
    "act_fun = 'relu' # activation function\n",
    "nfp = get_num_fingerprints(descriptor,elements) # number of fingerprints (or descriptors) from dataset\n",
    "arch = [10,10] # NN layers\n",
    "\n",
    "model = Network(elements, nfp, arch, act_fun)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Forces are used for training.\n",
      "Stresses are used for training.\n",
      "Scaling factors are computed using training dataset.\n",
      "Training dataset are standardized.\n",
      "Validation dataset are standardized.\n",
      "Training dataset will be shuffled during training.\n",
      "\n",
      "===> Epoch 1/50 - 0.287s/epoch\n",
      "     training_loss    - pe_loss: 31.190 - force_loss: 356.384 - stress_loss: 39990.078 - total_loss: 4386.582\n",
      "     validation_loss  - pe_loss: 32.375 - force_loss: 327.956 - stress_loss: 38602.905 - total_loss: 4220.621\n",
      "\n",
      "===> Epoch 2/50 - 0.255s/epoch\n",
      "     training_loss    - pe_loss: 29.309 - force_loss: 331.239 - stress_loss: 33570.114 - total_loss: 3717.558\n",
      "     validation_loss  - pe_loss: 28.926 - force_loss: 277.891 - stress_loss: 31481.089 - total_loss: 3454.926\n",
      "\n",
      "===> Epoch 3/50 - 0.243s/epoch\n",
      "     training_loss    - pe_loss: 19.546 - force_loss: 299.680 - stress_loss: 23171.993 - total_loss: 2636.426\n",
      "     validation_loss  - pe_loss: 26.623 - force_loss: 241.708 - stress_loss: 25293.141 - total_loss: 2797.644\n",
      "\n",
      "===> Epoch 4/50 - 0.267s/epoch\n",
      "     training_loss    - pe_loss: 17.683 - force_loss: 256.099 - stress_loss: 22112.295 - total_loss: 2485.011\n",
      "     validation_loss  - pe_loss: 24.879 - force_loss: 226.670 - stress_loss: 20505.446 - total_loss: 2302.093\n",
      "\n",
      "===> Epoch 5/50 - 0.206s/epoch\n",
      "     training_loss    - pe_loss: 19.918 - force_loss: 292.258 - stress_loss: 19392.847 - total_loss: 2251.461\n",
      "     validation_loss  - pe_loss: 23.761 - force_loss: 227.353 - stress_loss: 14071.244 - total_loss: 1658.238\n",
      "\n",
      "===> Epoch 6/50 - 0.217s/epoch\n",
      "     training_loss    - pe_loss: 18.440 - force_loss: 294.655 - stress_loss: 18934.791 - total_loss: 2206.574\n",
      "     validation_loss  - pe_loss: 23.181 - force_loss: 230.540 - stress_loss: 10633.239 - total_loss: 1317.044\n",
      "\n",
      "===> Epoch 7/50 - 0.250s/epoch\n",
      "     training_loss    - pe_loss: 27.071 - force_loss: 287.616 - stress_loss: 17824.723 - total_loss: 2097.160\n",
      "     validation_loss  - pe_loss: 22.483 - force_loss: 233.838 - stress_loss: 10535.882 - total_loss: 1309.910\n",
      "\n",
      "===> Epoch 8/50 - 0.213s/epoch\n",
      "     training_loss    - pe_loss: 19.139 - force_loss: 262.362 - stress_loss: 17680.789 - total_loss: 2049.580\n",
      "     validation_loss  - pe_loss: 21.706 - force_loss: 232.758 - stress_loss: 10097.616 - total_loss: 1264.227\n",
      "\n",
      "===> Epoch 9/50 - 0.247s/epoch\n",
      "     training_loss    - pe_loss: 16.393 - force_loss: 290.298 - stress_loss: 19667.135 - total_loss: 2273.404\n",
      "     validation_loss  - pe_loss: 20.813 - force_loss: 250.224 - stress_loss: 12001.393 - total_loss: 1471.177\n",
      "\n",
      "===> Epoch 10/50 - 0.211s/epoch\n",
      "     training_loss    - pe_loss: 19.249 - force_loss: 265.476 - stress_loss: 15017.614 - total_loss: 1786.486\n",
      "     validation_loss  - pe_loss: 19.828 - force_loss: 247.004 - stress_loss: 13313.849 - total_loss: 1598.217\n",
      "\n",
      "===> Epoch 11/50 - 0.247s/epoch\n",
      "     training_loss    - pe_loss: 17.911 - force_loss: 275.333 - stress_loss: 15653.270 - total_loss: 1858.571\n",
      "     validation_loss  - pe_loss: 18.833 - force_loss: 240.327 - stress_loss: 13014.488 - total_loss: 1560.609\n",
      "\n",
      "===> Epoch 12/50 - 0.239s/epoch\n",
      "     training_loss    - pe_loss: 17.082 - force_loss: 217.358 - stress_loss: 11507.945 - total_loss: 1385.234\n",
      "     validation_loss  - pe_loss: 17.622 - force_loss: 221.982 - stress_loss: 11865.776 - total_loss: 1426.182\n",
      "\n",
      "===> Epoch 13/50 - 0.211s/epoch\n",
      "     training_loss    - pe_loss: 15.717 - force_loss: 214.068 - stress_loss: 13056.794 - total_loss: 1535.465\n",
      "     validation_loss  - pe_loss: 16.475 - force_loss: 230.185 - stress_loss: 12524.101 - total_loss: 1499.071\n",
      "\n",
      "===> Epoch 14/50 - 0.212s/epoch\n",
      "     training_loss    - pe_loss: 14.213 - force_loss: 265.868 - stress_loss: 18321.999 - total_loss: 2112.281\n",
      "     validation_loss  - pe_loss: 15.383 - force_loss: 223.565 - stress_loss: 12250.728 - total_loss: 1464.020\n",
      "\n",
      "===> Epoch 15/50 - 0.256s/epoch\n",
      "     training_loss    - pe_loss: 15.886 - force_loss: 223.352 - stress_loss: 12262.024 - total_loss: 1465.440\n",
      "     validation_loss  - pe_loss: 14.432 - force_loss: 212.618 - stress_loss: 13039.163 - total_loss: 1530.967\n",
      "\n",
      "===> Epoch 16/50 - 0.255s/epoch\n",
      "     training_loss    - pe_loss: 10.931 - force_loss: 230.472 - stress_loss: 14679.957 - total_loss: 1709.399\n",
      "     validation_loss  - pe_loss: 14.192 - force_loss: 206.057 - stress_loss: 13255.603 - total_loss: 1545.809\n",
      "\n",
      "===> Epoch 17/50 - 0.254s/epoch\n",
      "     training_loss    - pe_loss: 14.032 - force_loss: 214.895 - stress_loss: 12529.994 - total_loss: 1481.926\n",
      "     validation_loss  - pe_loss: 14.426 - force_loss: 217.929 - stress_loss: 11939.140 - total_loss: 1426.269\n",
      "\n",
      "===> Epoch 18/50 - 0.258s/epoch\n",
      "     training_loss    - pe_loss: 13.977 - force_loss: 199.781 - stress_loss: 10151.721 - total_loss: 1228.930\n",
      "     validation_loss  - pe_loss: 14.610 - force_loss: 208.643 - stress_loss: 9057.785 - total_loss: 1129.031\n",
      "\n",
      "===> Epoch 19/50 - 0.237s/epoch\n",
      "     training_loss    - pe_loss: 14.160 - force_loss: 190.762 - stress_loss: 9122.232 - total_loss: 1117.146\n",
      "     validation_loss  - pe_loss: 15.200 - force_loss: 214.031 - stress_loss: 7970.364 - total_loss: 1026.267\n",
      "\n",
      "===> Epoch 20/50 - 0.567s/epoch\n",
      "     training_loss    - pe_loss: 12.521 - force_loss: 187.481 - stress_loss: 8241.700 - total_loss: 1024.172\n",
      "     validation_loss  - pe_loss: 15.501 - force_loss: 207.533 - stress_loss: 7092.256 - total_loss: 932.260\n",
      "\n",
      "===> Epoch 21/50 - 0.235s/epoch\n",
      "     training_loss    - pe_loss: 12.524 - force_loss: 218.186 - stress_loss: 11744.765 - total_loss: 1405.186\n",
      "     validation_loss  - pe_loss: 15.301 - force_loss: 203.474 - stress_loss: 7162.949 - total_loss: 935.070\n",
      "\n",
      "===> Epoch 22/50 - 0.240s/epoch\n",
      "     training_loss    - pe_loss: 14.047 - force_loss: 213.436 - stress_loss: 10236.759 - total_loss: 1251.159\n",
      "     validation_loss  - pe_loss: 14.761 - force_loss: 199.951 - stress_loss: 7072.908 - total_loss: 922.003\n",
      "\n",
      "===> Epoch 23/50 - 0.208s/epoch\n",
      "     training_loss    - pe_loss: 9.837 - force_loss: 248.349 - stress_loss: 11092.987 - total_loss: 1367.484\n",
      "     validation_loss  - pe_loss: 14.310 - force_loss: 194.913 - stress_loss: 6726.697 - total_loss: 881.892\n",
      "\n",
      "===> Epoch 24/50 - 0.244s/epoch\n",
      "     training_loss    - pe_loss: 11.016 - force_loss: 184.585 - stress_loss: 8991.769 - total_loss: 1094.778\n",
      "     validation_loss  - pe_loss: 14.106 - force_loss: 187.860 - stress_loss: 6918.478 - total_loss: 893.814\n",
      "\n",
      "===> Epoch 25/50 - 0.211s/epoch\n",
      "     training_loss    - pe_loss: 15.097 - force_loss: 184.636 - stress_loss: 8829.165 - total_loss: 1082.650\n",
      "     validation_loss  - pe_loss: 13.765 - force_loss: 186.991 - stress_loss: 7435.635 - total_loss: 944.320\n",
      "\n",
      "===> Epoch 26/50 - 0.242s/epoch\n",
      "     training_loss    - pe_loss: 12.930 - force_loss: 203.051 - stress_loss: 9268.753 - total_loss: 1142.856\n",
      "     validation_loss  - pe_loss: 13.263 - force_loss: 178.193 - stress_loss: 6942.370 - total_loss: 885.693\n",
      "\n",
      "===> Epoch 27/50 - 0.266s/epoch\n",
      "     training_loss    - pe_loss: 10.075 - force_loss: 180.792 - stress_loss: 9319.836 - total_loss: 1122.851\n",
      "     validation_loss  - pe_loss: 12.740 - force_loss: 172.928 - stress_loss: 6968.925 - total_loss: 882.560\n",
      "\n",
      "===> Epoch 28/50 - 0.303s/epoch\n",
      "     training_loss    - pe_loss: 9.762 - force_loss: 177.376 - stress_loss: 9268.017 - total_loss: 1113.940\n",
      "     validation_loss  - pe_loss: 12.235 - force_loss: 166.857 - stress_loss: 7379.211 - total_loss: 917.013\n",
      "\n",
      "===> Epoch 29/50 - 0.252s/epoch\n",
      "     training_loss    - pe_loss: 10.697 - force_loss: 178.297 - stress_loss: 6721.237 - total_loss: 861.117\n",
      "     validation_loss  - pe_loss: 11.725 - force_loss: 160.584 - stress_loss: 6881.283 - total_loss: 860.437\n",
      "\n",
      "===> Epoch 30/50 - 0.197s/epoch\n",
      "     training_loss    - pe_loss: 12.150 - force_loss: 168.624 - stress_loss: 8436.945 - total_loss: 1024.468\n",
      "     validation_loss  - pe_loss: 11.199 - force_loss: 166.258 - stress_loss: 7778.549 - total_loss: 955.312\n",
      "\n",
      "===> Epoch 31/50 - 0.232s/epoch\n",
      "     training_loss    - pe_loss: 13.997 - force_loss: 164.493 - stress_loss: 6873.374 - total_loss: 865.827\n",
      "     validation_loss  - pe_loss: 10.699 - force_loss: 160.313 - stress_loss: 7252.115 - total_loss: 896.223\n",
      "\n",
      "===> Epoch 32/50 - 0.218s/epoch\n",
      "     training_loss    - pe_loss: 11.401 - force_loss: 160.173 - stress_loss: 7200.905 - total_loss: 891.664\n",
      "     validation_loss  - pe_loss: 10.229 - force_loss: 154.167 - stress_loss: 7178.661 - total_loss: 882.263\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===> Epoch 33/50 - 0.246s/epoch\n",
      "     training_loss    - pe_loss: 9.099 - force_loss: 184.592 - stress_loss: 5416.165 - total_loss: 735.307\n",
      "     validation_loss  - pe_loss: 9.814 - force_loss: 150.473 - stress_loss: 6569.242 - total_loss: 817.212\n",
      "\n",
      "===> Epoch 34/50 - 0.228s/epoch\n",
      "     training_loss    - pe_loss: 8.762 - force_loss: 160.241 - stress_loss: 5852.148 - total_loss: 754.218\n",
      "     validation_loss  - pe_loss: 9.374 - force_loss: 145.594 - stress_loss: 6108.906 - total_loss: 765.858\n",
      "\n",
      "===> Epoch 35/50 - 0.240s/epoch\n",
      "     training_loss    - pe_loss: 12.406 - force_loss: 159.299 - stress_loss: 8944.486 - total_loss: 1066.153\n",
      "     validation_loss  - pe_loss: 8.900 - force_loss: 141.091 - stress_loss: 5858.747 - total_loss: 735.865\n",
      "\n",
      "===> Epoch 36/50 - 0.213s/epoch\n",
      "     training_loss    - pe_loss: 9.840 - force_loss: 149.606 - stress_loss: 7718.608 - total_loss: 931.307\n",
      "     validation_loss  - pe_loss: 8.535 - force_loss: 131.647 - stress_loss: 5924.420 - total_loss: 732.624\n",
      "\n",
      "===> Epoch 37/50 - 0.245s/epoch\n",
      "     training_loss    - pe_loss: 8.595 - force_loss: 147.061 - stress_loss: 6082.042 - total_loss: 763.860\n",
      "     validation_loss  - pe_loss: 8.216 - force_loss: 132.321 - stress_loss: 5114.304 - total_loss: 651.967\n",
      "\n",
      "===> Epoch 38/50 - 0.223s/epoch\n",
      "     training_loss    - pe_loss: 6.672 - force_loss: 129.401 - stress_loss: 5902.543 - total_loss: 726.327\n",
      "     validation_loss  - pe_loss: 7.922 - force_loss: 139.757 - stress_loss: 6516.377 - total_loss: 799.317\n",
      "\n",
      "===> Epoch 39/50 - 0.263s/epoch\n",
      "     training_loss    - pe_loss: 7.650 - force_loss: 133.333 - stress_loss: 6576.271 - total_loss: 798.610\n",
      "     validation_loss  - pe_loss: 7.689 - force_loss: 137.120 - stress_loss: 6384.103 - total_loss: 783.219\n",
      "\n",
      "===> Epoch 40/50 - 0.237s/epoch\n",
      "     training_loss    - pe_loss: 6.882 - force_loss: 129.057 - stress_loss: 5120.016 - total_loss: 647.940\n",
      "     validation_loss  - pe_loss: 7.396 - force_loss: 136.088 - stress_loss: 7042.040 - total_loss: 847.688\n",
      "\n",
      "===> Epoch 41/50 - 0.206s/epoch\n",
      "     training_loss    - pe_loss: 6.412 - force_loss: 109.322 - stress_loss: 6859.636 - total_loss: 801.697\n",
      "     validation_loss  - pe_loss: 7.084 - force_loss: 130.286 - stress_loss: 5301.032 - total_loss: 667.474\n",
      "\n",
      "===> Epoch 42/50 - 0.233s/epoch\n",
      "     training_loss    - pe_loss: 8.292 - force_loss: 124.199 - stress_loss: 5566.544 - total_loss: 689.145\n",
      "     validation_loss  - pe_loss: 6.817 - force_loss: 125.865 - stress_loss: 5427.751 - total_loss: 675.457\n",
      "\n",
      "===> Epoch 43/50 - 0.241s/epoch\n",
      "     training_loss    - pe_loss: 7.371 - force_loss: 111.575 - stress_loss: 4294.890 - total_loss: 548.434\n",
      "     validation_loss  - pe_loss: 6.633 - force_loss: 117.875 - stress_loss: 5009.373 - total_loss: 625.446\n",
      "\n",
      "===> Epoch 44/50 - 0.222s/epoch\n",
      "     training_loss    - pe_loss: 6.747 - force_loss: 115.083 - stress_loss: 5000.162 - total_loss: 621.846\n",
      "     validation_loss  - pe_loss: 6.580 - force_loss: 115.144 - stress_loss: 4611.756 - total_loss: 582.899\n",
      "\n",
      "===> Epoch 45/50 - 0.239s/epoch\n",
      "     training_loss    - pe_loss: 6.703 - force_loss: 162.029 - stress_loss: 6026.879 - total_loss: 771.421\n",
      "     validation_loss  - pe_loss: 6.608 - force_loss: 113.569 - stress_loss: 4733.484 - total_loss: 593.526\n",
      "\n",
      "===> Epoch 46/50 - 0.236s/epoch\n",
      "     training_loss    - pe_loss: 6.483 - force_loss: 147.566 - stress_loss: 7066.115 - total_loss: 860.661\n",
      "     validation_loss  - pe_loss: 6.658 - force_loss: 114.772 - stress_loss: 3948.175 - total_loss: 516.247\n",
      "\n",
      "===> Epoch 47/50 - 0.223s/epoch\n",
      "     training_loss    - pe_loss: 9.509 - force_loss: 158.753 - stress_loss: 5859.098 - total_loss: 754.172\n",
      "     validation_loss  - pe_loss: 6.646 - force_loss: 108.553 - stress_loss: 3693.647 - total_loss: 484.564\n",
      "\n",
      "===> Epoch 48/50 - 0.221s/epoch\n",
      "     training_loss    - pe_loss: 8.387 - force_loss: 109.751 - stress_loss: 5225.576 - total_loss: 640.695\n",
      "     validation_loss  - pe_loss: 6.606 - force_loss: 100.762 - stress_loss: 3702.446 - total_loss: 477.613\n",
      "\n",
      "===> Epoch 49/50 - 0.251s/epoch\n",
      "     training_loss    - pe_loss: 6.458 - force_loss: 109.626 - stress_loss: 5447.562 - total_loss: 660.840\n",
      "     validation_loss  - pe_loss: 6.442 - force_loss: 100.152 - stress_loss: 3516.438 - total_loss: 458.238\n",
      "\n",
      "===> Epoch 50/50 - 0.231s/epoch\n",
      "     training_loss    - pe_loss: 8.750 - force_loss: 105.299 - stress_loss: 3383.525 - total_loss: 452.402\n",
      "     validation_loss  - pe_loss: 6.207 - force_loss: 94.140 - stress_loss: 4186.339 - total_loss: 518.981\n",
      "\n",
      "End of training, elapsed time:  00:00:12\n"
     ]
    }
   ],
   "source": [
    "# Train the model \n",
    "\n",
    "opt = 'Adam' # optimizer\n",
    "loss_fun = 'mae' # loss function\n",
    "scaling = 'std' # scaling the traning data with standardization\n",
    "lr = 0.01 # learning rate\n",
    "loss_weights = {'pe' : 1, 'force' : 1, 'stress': 0.1} # the weights in loss function\n",
    "\n",
    "model.train(train_dataset,val_dataset,\n",
    "            optimizer=opt, \n",
    "            loss_fun = loss_fun, \n",
    "            batch_size=30, \n",
    "            lr=lr, \n",
    "            epochs=50, \n",
    "            scaling=scaling, \n",
    "            loss_weights=loss_weights, \n",
    "            compute_all_loss=True, \n",
    "            shuffle=True, \n",
    "            append_loss=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the training loss\n",
    "\n",
    "model.plot_loss(start_epoch=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Evaluation loss is:\n",
      "        pe_loss:       8.3364e+00\n",
      "     force_loss:       9.8733e+01\n",
      "    stress_loss:       4.5703e+03\n",
      "     total_loss:       5.6410e+02\n",
      "The total loss is computed using the loss weights - pe: 1.00 - force: 1.00 - stress: 0.10\n"
     ]
    }
   ],
   "source": [
    "# Evaluate using the first 5 data in test dataset\n",
    "\n",
    "model.evaluate(test_dataset.take(5),return_prediction=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'pe': array([  4.98966932, -15.53297727, -21.9046552 , -29.27170519,\n",
       "        -26.450315  ]),\n",
       " 'force': array([[[ -75.06978466, -172.92211148,  229.67395629],\n",
       "         [  74.41928761,  -38.89317527, -129.60283767],\n",
       "         [ 291.48027163,   79.89212325,  108.91411508],\n",
       "         [-290.82977488,  131.92316363, -208.98523383]],\n",
       " \n",
       "        [[  -1.12284797,  -46.38003397,  -56.78587395],\n",
       "         [  37.89161011,   60.28494093,  -68.44166021],\n",
       "         [ -45.4923701 ,  -90.1904247 ,  132.87016702],\n",
       "         [   8.7236079 ,   76.2855178 ,   -7.64263283]],\n",
       " \n",
       "        [[-106.47649583,  -12.34103851,  377.1129094 ],\n",
       "         [  80.46425249,  217.20247428, -316.56428174],\n",
       "         [ 150.11888887, -261.81965841,  112.99040383],\n",
       "         [-124.10664501,   56.95822197, -173.53903144]],\n",
       " \n",
       "        [[-263.93606985,   84.63016038, -102.77425547],\n",
       "         [-179.60252309,   15.28498485,  -69.12516193],\n",
       "         [ 317.26096084,  -41.53343411,  -15.18248703],\n",
       "         [ 126.27763224,  -58.38171139,  187.08190441]],\n",
       " \n",
       "        [[ 254.3949484 ,  107.85178033,  304.67510746],\n",
       "         [ -52.12582621,   17.22681417, -255.89791842],\n",
       "         [-182.44827854,  -27.31303965,   94.44380832],\n",
       "         [ -19.82084327,  -97.76555481, -143.22099734]]]),\n",
       " 'stress': array([[-6.20371789e+03,  1.13627470e+01, -2.37123066e+01,\n",
       "          1.13627484e+01, -6.39380530e+03, -1.62881455e+00,\n",
       "         -2.37123045e+01, -1.62881573e+00,  1.20551161e+00],\n",
       "        [-8.50114487e+03,  1.84839986e+00,  2.08877457e-01,\n",
       "          1.84840042e+00, -8.65504055e+03, -4.07797059e-01,\n",
       "          2.08877844e-01, -4.07797449e-01, -4.92153189e+00],\n",
       "        [ 3.63854203e+04, -1.97266425e+01, -9.44855164e+00,\n",
       "         -1.97266432e+01,  3.63329050e+04,  2.73247795e+00,\n",
       "         -9.44855559e+00,  2.73248313e+00,  1.26945233e+01],\n",
       "        [-6.59333928e+03,  5.05246045e+01,  1.00709374e+01,\n",
       "          5.05246048e+01, -6.35512008e+03, -2.80925787e+00,\n",
       "          1.00709363e+01, -2.80925581e+00,  7.70519751e+00],\n",
       "        [ 4.09253355e+03,  3.87355845e+00, -4.18005947e+00,\n",
       "          3.87355820e+00,  4.34530492e+03, -1.30450751e+00,\n",
       "         -4.18006231e+00, -1.30450772e+00, -8.47212142e+00]])}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# prediction using the first 5 data in test dataset\n",
    "\n",
    "input_dict = get_input_dict(dataset.take(5))\n",
    "model.predict(input_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save/load model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**save the trained model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: example.tfdnn/assets\n",
      "Network signatures and descriptor are written to example.tfdnn/parameters for LAMMPS simulation.\n"
     ]
    }
   ],
   "source": [
    "# we re-write the descriptor here to empasize that it should be the same one defined above\n",
    "descriptor = {'name': 'acsf', \n",
    "              'cutoff': 6.5,\n",
    "              'etaG2':[0.01,0.05,0.1,0.5,1,5,10], \n",
    "              'etaG4': [0.01], \n",
    "              'zeta': [0.08,0.2,1.0,5.0,10.0,50.0,100.0],\n",
    "              'lambda': [1.0, -1.0]}\n",
    "\n",
    "save_dir = 'example.tfdnn'\n",
    "network.save(model,save_dir,descriptor=descriptor)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Load the trained model for continuous training and prediction**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Network has been inflated! self.built: True\n",
      "Forces are used for training.\n",
      "Stresses are used for training.\n",
      "Scaling factors are computed using training dataset.\n",
      "Training dataset are standardized.\n",
      "Validation dataset are standardized.\n",
      "Training dataset will be shuffled during training.\n",
      "\n",
      "===> Epoch 1/50 - 0.286s/epoch\n",
      "     training_loss    - pe_loss: 6.225 - force_loss: 107.097 - stress_loss: 5896.093 - total_loss: 702.930\n",
      "     validation_loss  - pe_loss: 5.153 - force_loss: 81.355 - stress_loss: 3708.718 - total_loss: 457.380\n",
      "\n",
      "===> Epoch 2/50 - 0.291s/epoch\n",
      "     training_loss    - pe_loss: 6.683 - force_loss: 85.014 - stress_loss: 3550.454 - total_loss: 446.742\n",
      "     validation_loss  - pe_loss: 6.016 - force_loss: 72.757 - stress_loss: 3547.104 - total_loss: 433.484\n",
      "\n",
      "===> Epoch 3/50 - 0.208s/epoch\n",
      "     training_loss    - pe_loss: 4.510 - force_loss: 71.721 - stress_loss: 3633.668 - total_loss: 439.598\n",
      "     validation_loss  - pe_loss: 6.587 - force_loss: 69.578 - stress_loss: 3321.606 - total_loss: 408.326\n",
      "\n",
      "===> Epoch 4/50 - 0.313s/epoch\n",
      "     training_loss    - pe_loss: 4.918 - force_loss: 72.456 - stress_loss: 3999.166 - total_loss: 477.291\n",
      "     validation_loss  - pe_loss: 7.014 - force_loss: 61.233 - stress_loss: 3738.862 - total_loss: 442.133\n",
      "\n",
      "===> Epoch 5/50 - 0.269s/epoch\n",
      "     training_loss    - pe_loss: 4.612 - force_loss: 68.645 - stress_loss: 3438.338 - total_loss: 417.091\n",
      "     validation_loss  - pe_loss: 7.047 - force_loss: 63.952 - stress_loss: 2774.278 - total_loss: 348.426\n",
      "\n",
      "===> Epoch 6/50 - 0.219s/epoch\n",
      "     training_loss    - pe_loss: 6.011 - force_loss: 61.255 - stress_loss: 3458.065 - total_loss: 413.073\n",
      "     validation_loss  - pe_loss: 6.830 - force_loss: 63.454 - stress_loss: 3492.039 - total_loss: 419.488\n",
      "\n",
      "===> Epoch 7/50 - 0.216s/epoch\n",
      "     training_loss    - pe_loss: 5.938 - force_loss: 64.808 - stress_loss: 3637.062 - total_loss: 434.452\n",
      "     validation_loss  - pe_loss: 6.210 - force_loss: 57.891 - stress_loss: 3011.305 - total_loss: 365.231\n",
      "\n",
      "===> Epoch 8/50 - 0.225s/epoch\n",
      "     training_loss    - pe_loss: 4.245 - force_loss: 76.674 - stress_loss: 2692.130 - total_loss: 350.132\n",
      "     validation_loss  - pe_loss: 5.141 - force_loss: 56.055 - stress_loss: 2189.114 - total_loss: 280.108\n",
      "\n",
      "===> Epoch 9/50 - 0.227s/epoch\n",
      "     training_loss    - pe_loss: 3.599 - force_loss: 51.576 - stress_loss: 2436.548 - total_loss: 298.829\n",
      "     validation_loss  - pe_loss: 4.396 - force_loss: 58.601 - stress_loss: 2538.165 - total_loss: 316.814\n",
      "\n",
      "===> Epoch 10/50 - 0.258s/epoch\n",
      "     training_loss    - pe_loss: 3.344 - force_loss: 46.560 - stress_loss: 2197.354 - total_loss: 269.640\n",
      "     validation_loss  - pe_loss: 4.293 - force_loss: 47.936 - stress_loss: 2566.444 - total_loss: 308.873\n",
      "\n",
      "===> Epoch 11/50 - 0.239s/epoch\n",
      "     training_loss    - pe_loss: 3.545 - force_loss: 47.598 - stress_loss: 2464.916 - total_loss: 297.635\n",
      "     validation_loss  - pe_loss: 4.367 - force_loss: 45.184 - stress_loss: 2279.156 - total_loss: 277.467\n",
      "\n",
      "===> Epoch 12/50 - 0.245s/epoch\n",
      "     training_loss    - pe_loss: 3.830 - force_loss: 45.659 - stress_loss: 2423.599 - total_loss: 291.849\n",
      "     validation_loss  - pe_loss: 4.177 - force_loss: 40.014 - stress_loss: 1761.722 - total_loss: 220.364\n",
      "\n",
      "===> Epoch 13/50 - 0.258s/epoch\n",
      "     training_loss    - pe_loss: 3.439 - force_loss: 55.449 - stress_loss: 2754.962 - total_loss: 334.384\n",
      "     validation_loss  - pe_loss: 3.599 - force_loss: 38.619 - stress_loss: 1264.196 - total_loss: 168.638\n",
      "\n",
      "===> Epoch 14/50 - 0.244s/epoch\n",
      "     training_loss    - pe_loss: 3.371 - force_loss: 43.619 - stress_loss: 1720.694 - total_loss: 219.059\n",
      "     validation_loss  - pe_loss: 3.033 - force_loss: 37.956 - stress_loss: 964.631 - total_loss: 137.452\n",
      "\n",
      "===> Epoch 15/50 - 0.216s/epoch\n",
      "     training_loss    - pe_loss: 2.820 - force_loss: 43.306 - stress_loss: 1832.361 - total_loss: 229.362\n",
      "     validation_loss  - pe_loss: 2.787 - force_loss: 41.718 - stress_loss: 1062.309 - total_loss: 150.736\n",
      "\n",
      "===> Epoch 16/50 - 0.260s/epoch\n",
      "     training_loss    - pe_loss: 2.301 - force_loss: 37.596 - stress_loss: 2066.742 - total_loss: 246.571\n",
      "     validation_loss  - pe_loss: 2.631 - force_loss: 34.549 - stress_loss: 942.406 - total_loss: 131.421\n",
      "\n",
      "===> Epoch 17/50 - 0.205s/epoch\n",
      "     training_loss    - pe_loss: 2.490 - force_loss: 33.295 - stress_loss: 1390.026 - total_loss: 174.788\n",
      "     validation_loss  - pe_loss: 2.603 - force_loss: 28.231 - stress_loss: 1015.503 - total_loss: 132.384\n",
      "\n",
      "===> Epoch 18/50 - 0.223s/epoch\n",
      "     training_loss    - pe_loss: 2.607 - force_loss: 40.722 - stress_loss: 1264.683 - total_loss: 169.797\n",
      "     validation_loss  - pe_loss: 2.482 - force_loss: 28.160 - stress_loss: 1063.177 - total_loss: 136.959\n",
      "\n",
      "===> Epoch 19/50 - 0.243s/epoch\n",
      "     training_loss    - pe_loss: 2.484 - force_loss: 36.703 - stress_loss: 1246.367 - total_loss: 163.824\n",
      "     validation_loss  - pe_loss: 2.171 - force_loss: 25.673 - stress_loss: 646.808 - total_loss: 92.525\n",
      "\n",
      "===> Epoch 20/50 - 0.245s/epoch\n",
      "     training_loss    - pe_loss: 2.196 - force_loss: 37.554 - stress_loss: 1010.980 - total_loss: 140.848\n",
      "     validation_loss  - pe_loss: 2.050 - force_loss: 27.034 - stress_loss: 989.504 - total_loss: 128.035\n",
      "\n",
      "===> Epoch 21/50 - 0.313s/epoch\n",
      "     training_loss    - pe_loss: 1.950 - force_loss: 40.177 - stress_loss: 1302.699 - total_loss: 172.397\n",
      "     validation_loss  - pe_loss: 2.259 - force_loss: 28.655 - stress_loss: 923.259 - total_loss: 123.240\n",
      "\n",
      "===> Epoch 22/50 - 0.273s/epoch\n",
      "     training_loss    - pe_loss: 1.939 - force_loss: 33.098 - stress_loss: 920.506 - total_loss: 127.088\n",
      "     validation_loss  - pe_loss: 2.598 - force_loss: 27.771 - stress_loss: 920.784 - total_loss: 122.447\n",
      "\n",
      "===> Epoch 23/50 - 0.273s/epoch\n",
      "     training_loss    - pe_loss: 2.114 - force_loss: 30.801 - stress_loss: 1053.435 - total_loss: 138.258\n",
      "     validation_loss  - pe_loss: 2.808 - force_loss: 23.787 - stress_loss: 796.039 - total_loss: 106.198\n",
      "\n",
      "===> Epoch 24/50 - 0.284s/epoch\n",
      "     training_loss    - pe_loss: 2.508 - force_loss: 28.795 - stress_loss: 798.401 - total_loss: 111.143\n",
      "     validation_loss  - pe_loss: 2.946 - force_loss: 23.228 - stress_loss: 654.580 - total_loss: 91.632\n",
      "\n",
      "===> Epoch 25/50 - 0.224s/epoch\n",
      "     training_loss    - pe_loss: 2.822 - force_loss: 28.372 - stress_loss: 639.680 - total_loss: 95.162\n",
      "     validation_loss  - pe_loss: 2.937 - force_loss: 22.818 - stress_loss: 562.909 - total_loss: 82.047\n",
      "\n",
      "===> Epoch 26/50 - 0.268s/epoch\n",
      "     training_loss    - pe_loss: 2.591 - force_loss: 28.649 - stress_loss: 561.209 - total_loss: 87.362\n",
      "     validation_loss  - pe_loss: 2.668 - force_loss: 23.794 - stress_loss: 591.417 - total_loss: 85.604\n",
      "\n",
      "===> Epoch 27/50 - 0.254s/epoch\n",
      "     training_loss    - pe_loss: 2.355 - force_loss: 23.397 - stress_loss: 483.879 - total_loss: 74.140\n",
      "     validation_loss  - pe_loss: 2.109 - force_loss: 24.381 - stress_loss: 475.684 - total_loss: 74.059\n",
      "\n",
      "===> Epoch 28/50 - 0.254s/epoch\n",
      "     training_loss    - pe_loss: 2.215 - force_loss: 24.191 - stress_loss: 349.811 - total_loss: 61.387\n",
      "     validation_loss  - pe_loss: 1.648 - force_loss: 23.037 - stress_loss: 360.378 - total_loss: 60.722\n",
      "\n",
      "===> Epoch 29/50 - 0.313s/epoch\n",
      "     training_loss    - pe_loss: 1.653 - force_loss: 26.394 - stress_loss: 387.579 - total_loss: 66.805\n",
      "     validation_loss  - pe_loss: 1.442 - force_loss: 18.368 - stress_loss: 334.823 - total_loss: 53.291\n",
      "\n",
      "===> Epoch 30/50 - 0.303s/epoch\n",
      "     training_loss    - pe_loss: 1.154 - force_loss: 20.671 - stress_loss: 327.221 - total_loss: 54.548\n",
      "     validation_loss  - pe_loss: 1.302 - force_loss: 16.569 - stress_loss: 340.470 - total_loss: 51.917\n",
      "\n",
      "===> Epoch 31/50 - 0.226s/epoch\n",
      "     training_loss    - pe_loss: 1.310 - force_loss: 22.966 - stress_loss: 326.334 - total_loss: 56.910\n",
      "     validation_loss  - pe_loss: 1.202 - force_loss: 16.213 - stress_loss: 318.304 - total_loss: 49.245\n",
      "\n",
      "===> Epoch 32/50 - 0.233s/epoch\n",
      "     training_loss    - pe_loss: 1.443 - force_loss: 20.132 - stress_loss: 277.640 - total_loss: 49.338\n",
      "     validation_loss  - pe_loss: 1.136 - force_loss: 16.502 - stress_loss: 280.441 - total_loss: 45.682\n",
      "\n",
      "===> Epoch 33/50 - 0.254s/epoch\n",
      "     training_loss    - pe_loss: 1.371 - force_loss: 20.072 - stress_loss: 294.634 - total_loss: 50.907\n",
      "     validation_loss  - pe_loss: 1.115 - force_loss: 15.626 - stress_loss: 218.226 - total_loss: 38.564\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===> Epoch 34/50 - 0.234s/epoch\n",
      "     training_loss    - pe_loss: 1.158 - force_loss: 18.625 - stress_loss: 217.164 - total_loss: 41.499\n",
      "     validation_loss  - pe_loss: 1.145 - force_loss: 15.217 - stress_loss: 211.943 - total_loss: 37.557\n",
      "\n",
      "===> Epoch 35/50 - 0.262s/epoch\n",
      "     training_loss    - pe_loss: 0.966 - force_loss: 16.210 - stress_loss: 171.115 - total_loss: 34.287\n",
      "     validation_loss  - pe_loss: 1.496 - force_loss: 16.229 - stress_loss: 217.624 - total_loss: 39.487\n",
      "\n",
      "===> Epoch 36/50 - 0.207s/epoch\n",
      "     training_loss    - pe_loss: 1.585 - force_loss: 15.893 - stress_loss: 223.773 - total_loss: 39.856\n",
      "     validation_loss  - pe_loss: 1.950 - force_loss: 15.536 - stress_loss: 227.235 - total_loss: 40.210\n",
      "\n",
      "===> Epoch 37/50 - 0.226s/epoch\n",
      "     training_loss    - pe_loss: 2.034 - force_loss: 15.639 - stress_loss: 162.887 - total_loss: 33.961\n",
      "     validation_loss  - pe_loss: 1.884 - force_loss: 14.791 - stress_loss: 193.579 - total_loss: 36.033\n",
      "\n",
      "===> Epoch 38/50 - 0.251s/epoch\n",
      "     training_loss    - pe_loss: 1.547 - force_loss: 16.658 - stress_loss: 135.980 - total_loss: 31.803\n",
      "     validation_loss  - pe_loss: 1.340 - force_loss: 14.072 - stress_loss: 157.810 - total_loss: 31.193\n",
      "\n",
      "===> Epoch 39/50 - 0.244s/epoch\n",
      "     training_loss    - pe_loss: 0.891 - force_loss: 13.698 - stress_loss: 172.006 - total_loss: 31.789\n",
      "     validation_loss  - pe_loss: 1.055 - force_loss: 13.231 - stress_loss: 140.377 - total_loss: 28.324\n",
      "\n",
      "===> Epoch 40/50 - 0.219s/epoch\n",
      "     training_loss    - pe_loss: 1.238 - force_loss: 15.531 - stress_loss: 200.949 - total_loss: 36.864\n",
      "     validation_loss  - pe_loss: 1.002 - force_loss: 12.882 - stress_loss: 101.479 - total_loss: 24.033\n",
      "\n",
      "===> Epoch 41/50 - 0.223s/epoch\n",
      "     training_loss    - pe_loss: 0.755 - force_loss: 13.338 - stress_loss: 130.370 - total_loss: 27.130\n",
      "     validation_loss  - pe_loss: 0.930 - force_loss: 12.227 - stress_loss: 97.618 - total_loss: 22.919\n",
      "\n",
      "===> Epoch 42/50 - 0.226s/epoch\n",
      "     training_loss    - pe_loss: 0.833 - force_loss: 14.639 - stress_loss: 129.654 - total_loss: 28.437\n",
      "     validation_loss  - pe_loss: 0.882 - force_loss: 12.062 - stress_loss: 80.547 - total_loss: 20.998\n",
      "\n",
      "===> Epoch 43/50 - 0.246s/epoch\n",
      "     training_loss    - pe_loss: 1.131 - force_loss: 13.162 - stress_loss: 110.646 - total_loss: 25.357\n",
      "     validation_loss  - pe_loss: 0.868 - force_loss: 13.379 - stress_loss: 71.995 - total_loss: 21.447\n",
      "\n",
      "===> Epoch 44/50 - 0.250s/epoch\n",
      "     training_loss    - pe_loss: 0.952 - force_loss: 13.554 - stress_loss: 79.879 - total_loss: 22.494\n",
      "     validation_loss  - pe_loss: 0.932 - force_loss: 14.400 - stress_loss: 78.787 - total_loss: 23.211\n",
      "\n",
      "===> Epoch 45/50 - 0.242s/epoch\n",
      "     training_loss    - pe_loss: 0.758 - force_loss: 12.906 - stress_loss: 117.809 - total_loss: 25.445\n",
      "     validation_loss  - pe_loss: 0.988 - force_loss: 14.545 - stress_loss: 90.461 - total_loss: 24.579\n",
      "\n",
      "===> Epoch 46/50 - 0.186s/epoch\n",
      "     training_loss    - pe_loss: 0.701 - force_loss: 13.074 - stress_loss: 74.397 - total_loss: 21.215\n",
      "     validation_loss  - pe_loss: 0.977 - force_loss: 12.150 - stress_loss: 55.447 - total_loss: 18.672\n",
      "\n",
      "===> Epoch 47/50 - 0.242s/epoch\n",
      "     training_loss    - pe_loss: 0.743 - force_loss: 14.006 - stress_loss: 51.716 - total_loss: 19.921\n",
      "     validation_loss  - pe_loss: 0.940 - force_loss: 11.309 - stress_loss: 76.686 - total_loss: 19.918\n",
      "\n",
      "===> Epoch 48/50 - 0.227s/epoch\n",
      "     training_loss    - pe_loss: 0.777 - force_loss: 12.847 - stress_loss: 64.845 - total_loss: 20.109\n",
      "     validation_loss  - pe_loss: 0.931 - force_loss: 10.905 - stress_loss: 75.985 - total_loss: 19.434\n",
      "\n",
      "===> Epoch 49/50 - 0.238s/epoch\n",
      "     training_loss    - pe_loss: 0.679 - force_loss: 12.772 - stress_loss: 54.894 - total_loss: 18.940\n",
      "     validation_loss  - pe_loss: 0.953 - force_loss: 10.538 - stress_loss: 63.520 - total_loss: 17.843\n",
      "\n",
      "===> Epoch 50/50 - 0.252s/epoch\n",
      "     training_loss    - pe_loss: 0.789 - force_loss: 11.463 - stress_loss: 58.099 - total_loss: 18.062\n",
      "     validation_loss  - pe_loss: 0.978 - force_loss: 10.639 - stress_loss: 46.868 - total_loss: 16.304\n",
      "\n",
      "End of training, elapsed time:  00:00:12\n"
     ]
    }
   ],
   "source": [
    "imported_model = network.load(save_dir)\n",
    "\n",
    "# Re-train the model \n",
    "loss_weights = {'pe' : 1, 'force' : 1, 'stress': 0.1}\n",
    "\n",
    "opt = 'Adam'\n",
    "loss_fun = 'mae'\n",
    "scaling = 'std'\n",
    "\n",
    "model.train(train_dataset, val_dataset, \n",
    "            optimizer=opt, \n",
    "            loss_fun = loss_fun, \n",
    "            batch_size=30, \n",
    "            lr=0.02, \n",
    "            epochs=50, \n",
    "            scaling=scaling, \n",
    "            loss_weights=loss_weights, \n",
    "            compute_all_loss=True, \n",
    "            shuffle=True, \n",
    "            append_loss=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.plot_loss(start_epoch=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Evaluation loss is:\n",
      "        pe_loss:       8.3364e+00\n",
      "     force_loss:       9.8733e+01\n",
      "    stress_loss:       4.5703e+03\n",
      "     total_loss:       5.6410e+02\n",
      "The total loss is computed using the loss weights - pe: 1.00 - force: 1.00 - stress: 0.10\n"
     ]
    }
   ],
   "source": [
    "imported_model.evaluate(test_dataset.take(5),return_prediction=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'pe': array([-35.19343363, -30.5018231 , -13.10211931, -40.24163573,\n",
       "        -34.76370812]),\n",
       " 'force': array([[[  62.99615578,  -13.02417413,   35.15568535],\n",
       "         [ 125.17564353,   56.29076835,    4.00421797],\n",
       "         [-120.00388   ,  -31.82880756,    4.69251285],\n",
       "         [ -68.16791941,  -11.43778665,  -43.85241614]],\n",
       " \n",
       "        [[ -10.02454801,  -75.68597036,  -84.857218  ],\n",
       "         [  76.18432679,   28.68981993,  116.42437951],\n",
       "         [ 104.90041358,   37.58853954,  216.42067892],\n",
       "         [-171.06019292,    9.40761135, -247.98784042]],\n",
       " \n",
       "        [[-187.60105157,   72.76025473,   95.32019806],\n",
       "         [ 115.33638883,  -71.88966868,  -41.80044277],\n",
       "         [  69.78280937,  152.38949361,   73.49249946],\n",
       "         [   2.48185339, -153.26007932, -127.01225473]],\n",
       " \n",
       "        [[ -78.6992566 , -152.26434996,  -34.44163215],\n",
       "         [ 217.72729506,   42.42075491,  330.17770106],\n",
       "         [  61.01884481, -136.88389852,   -9.8914536 ],\n",
       "         [-200.04688344,  246.72749333, -285.84461536]],\n",
       " \n",
       "        [[ 343.49811374,   -6.76913846,   74.74235097],\n",
       "         [-171.71298205,   96.87952106,  -14.96776701],\n",
       "         [-198.50906892, -136.20298223,   16.80680101],\n",
       "         [  26.7239375 ,   46.09259955,  -76.581385  ]]]),\n",
       " 'stress': array([[-1.96553956e+03, -3.84383326e+01,  5.27824788e+00,\n",
       "         -3.84383324e+01, -2.24581615e+03,  1.04064708e+00,\n",
       "          5.27824870e+00,  1.04064694e+00,  1.57429504e+00],\n",
       "        [-1.04472838e+04, -3.39653871e+01, -1.10222099e+01,\n",
       "         -3.39653872e+01, -1.05141466e+04,  3.61939603e-01,\n",
       "         -1.10222056e+01,  3.61936112e-01, -1.71367644e+01],\n",
       "        [ 4.39121460e+03,  4.09626905e+01, -1.07343792e+01,\n",
       "          4.09626917e+01,  4.26569843e+03, -5.34677434e+00,\n",
       "         -1.07343794e+01, -5.34677692e+00, -5.61152547e-01],\n",
       "        [ 3.66647618e+04,  3.74010913e+01,  1.02510058e+01,\n",
       "          3.74010908e+01,  3.65224082e+04,  9.64309282e-01,\n",
       "          1.02510072e+01,  9.64311133e-01,  1.47602297e+01],\n",
       "        [-1.49339234e+04, -1.26911479e+02, -8.38920551e+00,\n",
       "         -1.26911479e+02, -1.47875442e+04,  1.52792284e+00,\n",
       "         -8.38920762e+00,  1.52792340e+00, -2.73513556e+00]])}"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "input_dict = get_input_dict(test_dataset.take(5))\n",
    "imported_model.predict(input_dict)"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "hide_input": false,
  "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": "3.6.9"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "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
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}