notebooks/final_results.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from NeuralNetwork import NeuralNetwork\n",
    "from utils import get_data, onehot_encoding, grid_search, save_dict_to_file, load_dict_from_file, cross_validation\n",
    "from Layer import Layer, Input\n",
    "from functions import accuracy, MSE, MEE\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "DIR = '/Users/HP/Desktop/UNI/LM_1/MachineLearning/ML_prj/' # Directory of the package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_frac = 1/4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing training and blind test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "names = ['id', 'feature_1', 'feature_2', 'feature_3', 'feature_4', 'feature_5', 'feature_6', \n",
    "         'feature_7', 'feature_8', 'feature_9', 'feature_10', 'target_x', 'target_y','target_z']\n",
    "\n",
    "df = pd.read_csv(f\"{DIR}data/ML-CUP23-TR.csv\", names=names, comment='#')\n",
    "\n",
    "names_test = ['id', 'feature_1', 'feature_2', 'feature_3', 'feature_4', 'feature_5', 'feature_6', \n",
    "         'feature_7', 'feature_8', 'feature_9', 'feature_10']\n",
    "\n",
    "df_test = pd.read_csv(f\"{DIR}data/ML-CUP23-TS.csv\", names=names_test, comment='#')\n",
    "\n",
    "targets = ['target_x', 'target_y', 'target_z']\n",
    "features = list(set(names) - {'id', 'target_x', 'target_y', 'target_z'})\n",
    "\n",
    "df = df.sample(frac=1)\n",
    "\n",
    "train_test_div = round((1-test_frac) * df.shape[0])\n",
    "\n",
    "X_train, y_train = df[:train_test_div][features].to_numpy().T, df[:train_test_div][targets].to_numpy().T\n",
    "X_test, y_test = df[train_test_div:][features].to_numpy().T, df[train_test_div:][targets].to_numpy().T\n",
    "\n",
    "X_blind_test = df_test[features].to_numpy().T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Best parameter combination:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_params = {'eta': 0.001,\n",
    "            'lam': 3e-05,\n",
    "            'alpha': 0.9,\n",
    "            'epochs': 500,\n",
    "            'n_batch': 150,\n",
    "            'scale_eta_batchsize': None,\n",
    "            'dim_hidden': 70,\n",
    "            'hidden_act_func': 'tanh',\n",
    "            'dim_hidden2': 70,\n",
    "            'hidden_act_func2': 'tanh',\n",
    "            'use_opt': 0,\n",
    "            'loss': 'MSE',\n",
    "            'output_act_func': 'lin',\n",
    "            'nest': True,\n",
    "            }\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "history_cv = cross_validation(X_train, y_train, folds=5, metrics=[MEE], params=best_params.copy(), callbacks={'early_stopping':None, 'reduce_eta':None})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'train_loss_mean': 0.25480430266237875, 'train_loss_std': 0.0066180402759764365, 'val_loss_mean': 0.5283306144367964, 'val_loss_std': 0.15485615040433304, 'train_MEE_mean': 0.692517436479362, 'train_MEE_std': 0.005851215083652159, 'val_MEE_mean': 0.9319407548517976, 'val_MEE_std': 0.06531931061399407}\n"
     ]
    }
   ],
   "source": [
    "print(history_cv)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Error evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = best_params.copy()\n",
    "params.pop('scale_eta_batchsize')\n",
    "input_layer = Input(X_train.shape[0])\n",
    "hidden_layer = Layer(input_layer, params.pop('dim_hidden'), params.pop('hidden_act_func'))\n",
    "o = 2\n",
    "while True:\n",
    "    if f'dim_hidden{o}' in params.keys():\n",
    "        hidden_layer = Layer(hidden_layer, params.pop(f'dim_hidden{o}'), params.pop(f'hidden_act_func{o}'))\n",
    "        o += 1\n",
    "    else: break\n",
    "output_layer = Layer(hidden_layer, 3, params.pop('output_act_func'))\n",
    "\n",
    "model = NeuralNetwork(input_layer, output_layer, params.pop('loss'), metrics=[MEE])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0: train_loss: 4.046e+02, test_loss: 4.103e+02; train_MEE: 2.931e+01, test_MEE: 2.963e+01  lr : 0.002\n",
      "Epoch 1: train_loss: 7.196e+01, test_loss: 7.594e+01; train_MEE: 1.245e+01, test_MEE: 1.282e+01  lr : 0.002\n",
      "Epoch 2: train_loss: 2.901e+01, test_loss: 2.941e+01; train_MEE: 8.362e+00, test_MEE: 8.413e+00  lr : 0.002\n",
      "Epoch 3: train_loss: 1.723e+01, test_loss: 1.697e+01; train_MEE: 6.156e+00, test_MEE: 6.278e+00  lr : 0.002\n",
      "Epoch 4: train_loss: 1.075e+01, test_loss: 9.532e+00; train_MEE: 4.746e+00, test_MEE: 4.706e+00  lr : 0.002\n",
      "Epoch 5: train_loss: 7.632e+00, test_loss: 6.328e+00; train_MEE: 3.935e+00, test_MEE: 3.843e+00  lr : 0.002\n",
      "Epoch 6: train_loss: 5.815e+00, test_loss: 5.159e+00; train_MEE: 3.411e+00, test_MEE: 3.431e+00  lr : 0.002\n",
      "Epoch 7: train_loss: 4.645e+00, test_loss: 4.326e+00; train_MEE: 2.964e+00, test_MEE: 3.092e+00  lr : 0.002\n",
      "Epoch 8: train_loss: 3.893e+00, test_loss: 3.646e+00; train_MEE: 2.668e+00, test_MEE: 2.838e+00  lr : 0.002\n",
      "Epoch 9: train_loss: 3.398e+00, test_loss: 2.953e+00; train_MEE: 2.452e+00, test_MEE: 2.516e+00  lr : 0.002\n",
      "Epoch 10: train_loss: 2.954e+00, test_loss: 2.483e+00; train_MEE: 2.254e+00, test_MEE: 2.292e+00  lr : 0.002\n",
      "Epoch 11: train_loss: 2.616e+00, test_loss: 2.218e+00; train_MEE: 2.105e+00, test_MEE: 2.158e+00  lr : 0.002\n",
      "Epoch 12: train_loss: 2.379e+00, test_loss: 2.086e+00; train_MEE: 1.993e+00, test_MEE: 2.077e+00  lr : 0.002\n",
      "Epoch 13: train_loss: 2.165e+00, test_loss: 1.916e+00; train_MEE: 1.892e+00, test_MEE: 1.983e+00  lr : 0.002\n",
      "Epoch 14: train_loss: 1.984e+00, test_loss: 1.775e+00; train_MEE: 1.813e+00, test_MEE: 1.906e+00  lr : 0.002\n",
      "Epoch 15: train_loss: 1.828e+00, test_loss: 1.661e+00; train_MEE: 1.732e+00, test_MEE: 1.833e+00  lr : 0.002\n",
      "Epoch 16: train_loss: 1.688e+00, test_loss: 1.569e+00; train_MEE: 1.667e+00, test_MEE: 1.782e+00  lr : 0.002\n",
      "Epoch 17: train_loss: 1.574e+00, test_loss: 1.468e+00; train_MEE: 1.611e+00, test_MEE: 1.730e+00  lr : 0.002\n",
      "Epoch 18: train_loss: 1.459e+00, test_loss: 1.415e+00; train_MEE: 1.553e+00, test_MEE: 1.702e+00  lr : 0.002\n",
      "Epoch 19: train_loss: 1.360e+00, test_loss: 1.322e+00; train_MEE: 1.503e+00, test_MEE: 1.650e+00  lr : 0.002\n",
      "Epoch 20: train_loss: 1.268e+00, test_loss: 1.266e+00; train_MEE: 1.462e+00, test_MEE: 1.623e+00  lr : 0.002\n",
      "Epoch 21: train_loss: 1.192e+00, test_loss: 1.220e+00; train_MEE: 1.420e+00, test_MEE: 1.595e+00  lr : 0.002\n",
      "Epoch 22: train_loss: 1.118e+00, test_loss: 1.158e+00; train_MEE: 1.378e+00, test_MEE: 1.556e+00  lr : 0.002\n",
      "Epoch 23: train_loss: 1.050e+00, test_loss: 1.130e+00; train_MEE: 1.344e+00, test_MEE: 1.542e+00  lr : 0.002\n",
      "Epoch 24: train_loss: 9.848e-01, test_loss: 1.084e+00; train_MEE: 1.310e+00, test_MEE: 1.517e+00  lr : 0.002\n",
      "Epoch 25: train_loss: 9.343e-01, test_loss: 1.057e+00; train_MEE: 1.277e+00, test_MEE: 1.498e+00  lr : 0.002\n",
      "Epoch 26: train_loss: 8.899e-01, test_loss: 1.041e+00; train_MEE: 1.255e+00, test_MEE: 1.486e+00  lr : 0.002\n",
      "Epoch 27: train_loss: 8.478e-01, test_loss: 9.893e-01; train_MEE: 1.220e+00, test_MEE: 1.449e+00  lr : 0.002\n",
      "Epoch 28: train_loss: 8.094e-01, test_loss: 9.666e-01; train_MEE: 1.198e+00, test_MEE: 1.439e+00  lr : 0.002\n",
      "Epoch 29: train_loss: 7.752e-01, test_loss: 9.422e-01; train_MEE: 1.173e+00, test_MEE: 1.415e+00  lr : 0.002\n",
      "Epoch 30: train_loss: 7.425e-01, test_loss: 9.106e-01; train_MEE: 1.150e+00, test_MEE: 1.394e+00  lr : 0.002\n",
      "Epoch 31: train_loss: 7.085e-01, test_loss: 8.865e-01; train_MEE: 1.127e+00, test_MEE: 1.373e+00  lr : 0.002\n",
      "Epoch 32: train_loss: 6.839e-01, test_loss: 8.602e-01; train_MEE: 1.112e+00, test_MEE: 1.351e+00  lr : 0.002\n",
      "Epoch 33: train_loss: 6.487e-01, test_loss: 8.367e-01; train_MEE: 1.087e+00, test_MEE: 1.338e+00  lr : 0.002\n",
      "Epoch 34: train_loss: 6.209e-01, test_loss: 8.154e-01; train_MEE: 1.074e+00, test_MEE: 1.319e+00  lr : 0.002\n",
      "Epoch 35: train_loss: 5.993e-01, test_loss: 7.991e-01; train_MEE: 1.057e+00, test_MEE: 1.303e+00  lr : 0.002\n",
      "Epoch 36: train_loss: 5.762e-01, test_loss: 7.736e-01; train_MEE: 1.042e+00, test_MEE: 1.280e+00  lr : 0.002\n",
      "Epoch 37: train_loss: 5.547e-01, test_loss: 7.638e-01; train_MEE: 1.029e+00, test_MEE: 1.269e+00  lr : 0.002\n",
      "Epoch 38: train_loss: 5.375e-01, test_loss: 7.527e-01; train_MEE: 1.019e+00, test_MEE: 1.262e+00  lr : 0.002\n",
      "Epoch 39: train_loss: 5.236e-01, test_loss: 7.346e-01; train_MEE: 1.002e+00, test_MEE: 1.241e+00  lr : 0.002\n",
      "Epoch 40: train_loss: 5.054e-01, test_loss: 7.288e-01; train_MEE: 9.902e-01, test_MEE: 1.237e+00  lr : 0.002\n",
      "Epoch 41: train_loss: 4.922e-01, test_loss: 7.151e-01; train_MEE: 9.745e-01, test_MEE: 1.218e+00  lr : 0.002\n",
      "Epoch 42: train_loss: 4.799e-01, test_loss: 7.154e-01; train_MEE: 9.671e-01, test_MEE: 1.218e+00  lr : 0.002\n",
      "Epoch 43: train_loss: 4.692e-01, test_loss: 6.984e-01; train_MEE: 9.563e-01, test_MEE: 1.206e+00  lr : 0.002\n",
      "Epoch 44: train_loss: 4.568e-01, test_loss: 6.741e-01; train_MEE: 9.447e-01, test_MEE: 1.179e+00  lr : 0.002\n",
      "Epoch 45: train_loss: 4.450e-01, test_loss: 6.725e-01; train_MEE: 9.337e-01, test_MEE: 1.176e+00  lr : 0.002\n",
      "Epoch 46: train_loss: 4.357e-01, test_loss: 6.617e-01; train_MEE: 9.261e-01, test_MEE: 1.167e+00  lr : 0.002\n",
      "Epoch 47: train_loss: 4.292e-01, test_loss: 6.569e-01; train_MEE: 9.233e-01, test_MEE: 1.164e+00  lr : 0.002\n",
      "Epoch 48: train_loss: 4.196e-01, test_loss: 6.457e-01; train_MEE: 9.113e-01, test_MEE: 1.149e+00  lr : 0.002\n",
      "Epoch 49: train_loss: 4.126e-01, test_loss: 6.481e-01; train_MEE: 9.061e-01, test_MEE: 1.150e+00  lr : 0.002\n",
      "Epoch 50: train_loss: 4.053e-01, test_loss: 6.452e-01; train_MEE: 8.969e-01, test_MEE: 1.148e+00  lr : 0.002\n",
      "Epoch 51: train_loss: 3.968e-01, test_loss: 6.232e-01; train_MEE: 8.886e-01, test_MEE: 1.124e+00  lr : 0.002\n",
      "Epoch 52: train_loss: 3.896e-01, test_loss: 6.198e-01; train_MEE: 8.826e-01, test_MEE: 1.121e+00  lr : 0.002\n",
      "Epoch 53: train_loss: 3.829e-01, test_loss: 6.161e-01; train_MEE: 8.700e-01, test_MEE: 1.117e+00  lr : 0.002\n",
      "Epoch 54: train_loss: 3.754e-01, test_loss: 6.105e-01; train_MEE: 8.635e-01, test_MEE: 1.107e+00  lr : 0.002\n",
      "Epoch 55: train_loss: 3.690e-01, test_loss: 6.082e-01; train_MEE: 8.576e-01, test_MEE: 1.103e+00  lr : 0.002\n",
      "Epoch 56: train_loss: 3.621e-01, test_loss: 6.020e-01; train_MEE: 8.486e-01, test_MEE: 1.094e+00  lr : 0.002\n",
      "Epoch 57: train_loss: 3.595e-01, test_loss: 5.937e-01; train_MEE: 8.461e-01, test_MEE: 1.087e+00  lr : 0.002\n",
      "Epoch 58: train_loss: 3.529e-01, test_loss: 5.936e-01; train_MEE: 8.391e-01, test_MEE: 1.084e+00  lr : 0.002\n",
      "Epoch 59: train_loss: 3.480e-01, test_loss: 5.851e-01; train_MEE: 8.323e-01, test_MEE: 1.075e+00  lr : 0.002\n",
      "Epoch 60: train_loss: 3.426e-01, test_loss: 5.824e-01; train_MEE: 8.258e-01, test_MEE: 1.071e+00  lr : 0.002\n",
      "Epoch 61: train_loss: 3.375e-01, test_loss: 5.803e-01; train_MEE: 8.233e-01, test_MEE: 1.067e+00  lr : 0.002\n",
      "Epoch 62: train_loss: 3.361e-01, test_loss: 5.719e-01; train_MEE: 8.181e-01, test_MEE: 1.060e+00  lr : 0.002\n",
      "Epoch 63: train_loss: 3.284e-01, test_loss: 5.687e-01; train_MEE: 8.105e-01, test_MEE: 1.056e+00  lr : 0.002\n",
      "Epoch 64: train_loss: 3.257e-01, test_loss: 5.674e-01; train_MEE: 8.057e-01, test_MEE: 1.053e+00  lr : 0.002\n",
      "Epoch 65: train_loss: 3.206e-01, test_loss: 5.581e-01; train_MEE: 8.012e-01, test_MEE: 1.042e+00  lr : 0.002\n",
      "Epoch 66: train_loss: 3.172e-01, test_loss: 5.523e-01; train_MEE: 7.972e-01, test_MEE: 1.038e+00  lr : 0.002\n",
      "Epoch 67: train_loss: 3.134e-01, test_loss: 5.505e-01; train_MEE: 7.914e-01, test_MEE: 1.034e+00  lr : 0.002\n",
      "Epoch 68: train_loss: 3.118e-01, test_loss: 5.504e-01; train_MEE: 7.898e-01, test_MEE: 1.028e+00  lr : 0.002\n",
      "Epoch 69: train_loss: 3.073e-01, test_loss: 5.438e-01; train_MEE: 7.860e-01, test_MEE: 1.024e+00  lr : 0.002\n",
      "Epoch 70: train_loss: 3.045e-01, test_loss: 5.419e-01; train_MEE: 7.822e-01, test_MEE: 1.023e+00  lr : 0.002\n",
      "Epoch 71: train_loss: 2.992e-01, test_loss: 5.387e-01; train_MEE: 7.739e-01, test_MEE: 1.014e+00  lr : 0.002\n",
      "Epoch 72: train_loss: 2.954e-01, test_loss: 5.370e-01; train_MEE: 7.690e-01, test_MEE: 1.013e+00  lr : 0.002\n",
      "Epoch 73: train_loss: 2.937e-01, test_loss: 5.297e-01; train_MEE: 7.660e-01, test_MEE: 1.003e+00  lr : 0.002\n",
      "Epoch 74: train_loss: 2.962e-01, test_loss: 5.322e-01; train_MEE: 7.745e-01, test_MEE: 1.014e+00  lr : 0.002\n",
      "Epoch 75: train_loss: 2.989e-01, test_loss: 5.332e-01; train_MEE: 7.849e-01, test_MEE: 1.022e+00  lr : 0.002\n",
      "Epoch 76: train_loss: 2.953e-01, test_loss: 5.289e-01; train_MEE: 7.757e-01, test_MEE: 1.011e+00  lr : 0.002\n",
      "Epoch 77: train_loss: 2.890e-01, test_loss: 5.226e-01; train_MEE: 7.653e-01, test_MEE: 1.002e+00  lr : 0.002\n",
      "Epoch 78: train_loss: 2.825e-01, test_loss: 5.120e-01; train_MEE: 7.535e-01, test_MEE: 9.836e-01  lr : 0.002\n",
      "Epoch 79: train_loss: 2.768e-01, test_loss: 5.074e-01; train_MEE: 7.432e-01, test_MEE: 9.799e-01  lr : 0.002\n",
      "Epoch 80: train_loss: 2.749e-01, test_loss: 5.136e-01; train_MEE: 7.405e-01, test_MEE: 9.816e-01  lr : 0.002\n",
      "Epoch 81: train_loss: 2.721e-01, test_loss: 5.079e-01; train_MEE: 7.376e-01, test_MEE: 9.745e-01  lr : 0.002\n",
      "Epoch 82: train_loss: 2.708e-01, test_loss: 5.005e-01; train_MEE: 7.349e-01, test_MEE: 9.680e-01  lr : 0.002\n",
      "Epoch 83: train_loss: 2.685e-01, test_loss: 5.022e-01; train_MEE: 7.354e-01, test_MEE: 9.717e-01  lr : 0.002\n",
      "Epoch 84: train_loss: 2.712e-01, test_loss: 5.081e-01; train_MEE: 7.444e-01, test_MEE: 9.836e-01  lr : 0.002\n",
      "Epoch 85: train_loss: 2.666e-01, test_loss: 5.028e-01; train_MEE: 7.329e-01, test_MEE: 9.662e-01  lr : 0.002\n",
      "Epoch 86: train_loss: 2.608e-01, test_loss: 4.972e-01; train_MEE: 7.227e-01, test_MEE: 9.627e-01  lr : 0.002\n",
      "Epoch 87: train_loss: 2.604e-01, test_loss: 4.958e-01; train_MEE: 7.228e-01, test_MEE: 9.605e-01  lr : 0.002\n",
      "Epoch 88: train_loss: 2.560e-01, test_loss: 4.904e-01; train_MEE: 7.150e-01, test_MEE: 9.513e-01  lr : 0.002\n",
      "Epoch 89: train_loss: 2.558e-01, test_loss: 4.922e-01; train_MEE: 7.140e-01, test_MEE: 9.531e-01  lr : 0.002\n",
      "Epoch 90: train_loss: 2.563e-01, test_loss: 4.933e-01; train_MEE: 7.188e-01, test_MEE: 9.533e-01  lr : 0.002\n",
      "Epoch 91: train_loss: 2.530e-01, test_loss: 4.874e-01; train_MEE: 7.131e-01, test_MEE: 9.497e-01  lr : 0.002\n",
      "Epoch 92: train_loss: 2.491e-01, test_loss: 4.840e-01; train_MEE: 7.048e-01, test_MEE: 9.411e-01  lr : 0.002\n",
      "Epoch 93: train_loss: 2.494e-01, test_loss: 4.853e-01; train_MEE: 7.047e-01, test_MEE: 9.446e-01  lr : 0.002\n",
      "Epoch 94: train_loss: 2.504e-01, test_loss: 4.864e-01; train_MEE: 7.084e-01, test_MEE: 9.470e-01  lr : 0.002\n",
      "Epoch 95: train_loss: 2.521e-01, test_loss: 4.885e-01; train_MEE: 7.118e-01, test_MEE: 9.515e-01  lr : 0.002\n",
      "Epoch 96: train_loss: 2.499e-01, test_loss: 4.842e-01; train_MEE: 7.125e-01, test_MEE: 9.490e-01  lr : 0.002\n",
      "Epoch 97: train_loss: 2.603e-01, test_loss: 4.955e-01; train_MEE: 7.364e-01, test_MEE: 9.670e-01  lr : 0.002\n",
      "Epoch 98: train_loss: 2.685e-01, test_loss: 4.986e-01; train_MEE: 7.539e-01, test_MEE: 9.812e-01  lr : 0.002\n",
      "Epoch 99: train_loss: 2.471e-01, test_loss: 4.814e-01; train_MEE: 7.079e-01, test_MEE: 9.430e-01  lr : 0.002\n",
      "Epoch 100: train_loss: 2.389e-01, test_loss: 4.736e-01; train_MEE: 6.907e-01, test_MEE: 9.293e-01  lr : 0.002\n",
      "Epoch 101: train_loss: 2.359e-01, test_loss: 4.722e-01; train_MEE: 6.857e-01, test_MEE: 9.236e-01  lr : 0.002\n",
      "Epoch 102: train_loss: 2.351e-01, test_loss: 4.706e-01; train_MEE: 6.856e-01, test_MEE: 9.240e-01  lr : 0.002\n",
      "Epoch 103: train_loss: 2.327e-01, test_loss: 4.692e-01; train_MEE: 6.808e-01, test_MEE: 9.178e-01  lr : 0.002\n",
      "Epoch 104: train_loss: 2.309e-01, test_loss: 4.645e-01; train_MEE: 6.780e-01, test_MEE: 9.168e-01  lr : 0.002\n",
      "Epoch 105: train_loss: 2.304e-01, test_loss: 4.672e-01; train_MEE: 6.755e-01, test_MEE: 9.191e-01  lr : 0.002\n",
      "Epoch 106: train_loss: 2.285e-01, test_loss: 4.597e-01; train_MEE: 6.711e-01, test_MEE: 9.106e-01  lr : 0.002\n",
      "Epoch 107: train_loss: 2.297e-01, test_loss: 4.627e-01; train_MEE: 6.768e-01, test_MEE: 9.175e-01  lr : 0.002\n",
      "Epoch 108: train_loss: 2.305e-01, test_loss: 4.684e-01; train_MEE: 6.814e-01, test_MEE: 9.248e-01  lr : 0.002\n",
      "Epoch 109: train_loss: 2.326e-01, test_loss: 4.689e-01; train_MEE: 6.876e-01, test_MEE: 9.226e-01  lr : 0.002\n",
      "Epoch 110: train_loss: 2.253e-01, test_loss: 4.601e-01; train_MEE: 6.702e-01, test_MEE: 9.087e-01  lr : 0.002\n",
      "Epoch 111: train_loss: 2.269e-01, test_loss: 4.625e-01; train_MEE: 6.743e-01, test_MEE: 9.165e-01  lr : 0.002\n",
      "Epoch 112: train_loss: 2.236e-01, test_loss: 4.567e-01; train_MEE: 6.665e-01, test_MEE: 9.079e-01  lr : 0.002\n",
      "Epoch 113: train_loss: 2.218e-01, test_loss: 4.535e-01; train_MEE: 6.637e-01, test_MEE: 9.022e-01  lr : 0.002\n",
      "Epoch 114: train_loss: 2.198e-01, test_loss: 4.499e-01; train_MEE: 6.600e-01, test_MEE: 8.973e-01  lr : 0.002\n",
      "Epoch 115: train_loss: 2.187e-01, test_loss: 4.513e-01; train_MEE: 6.563e-01, test_MEE: 8.947e-01  lr : 0.002\n",
      "Epoch 116: train_loss: 2.180e-01, test_loss: 4.542e-01; train_MEE: 6.582e-01, test_MEE: 9.054e-01  lr : 0.002\n",
      "Epoch 117: train_loss: 2.188e-01, test_loss: 4.495e-01; train_MEE: 6.602e-01, test_MEE: 8.996e-01  lr : 0.002\n",
      "Epoch 118: train_loss: 2.161e-01, test_loss: 4.472e-01; train_MEE: 6.570e-01, test_MEE: 8.921e-01  lr : 0.002\n",
      "Epoch 119: train_loss: 2.178e-01, test_loss: 4.497e-01; train_MEE: 6.606e-01, test_MEE: 8.962e-01  lr : 0.002\n",
      "Epoch 120: train_loss: 2.143e-01, test_loss: 4.479e-01; train_MEE: 6.519e-01, test_MEE: 8.929e-01  lr : 0.002\n",
      "Epoch 121: train_loss: 2.130e-01, test_loss: 4.394e-01; train_MEE: 6.483e-01, test_MEE: 8.823e-01  lr : 0.002\n",
      "Epoch 122: train_loss: 2.110e-01, test_loss: 4.400e-01; train_MEE: 6.467e-01, test_MEE: 8.841e-01  lr : 0.002\n",
      "Epoch 123: train_loss: 2.121e-01, test_loss: 4.351e-01; train_MEE: 6.479e-01, test_MEE: 8.799e-01  lr : 0.002\n",
      "Epoch 124: train_loss: 2.140e-01, test_loss: 4.416e-01; train_MEE: 6.547e-01, test_MEE: 8.896e-01  lr : 0.002\n",
      "Epoch 125: train_loss: 2.163e-01, test_loss: 4.446e-01; train_MEE: 6.626e-01, test_MEE: 8.960e-01  lr : 0.002\n",
      "Epoch 126: train_loss: 2.102e-01, test_loss: 4.344e-01; train_MEE: 6.478e-01, test_MEE: 8.778e-01  lr : 0.002\n",
      "Epoch 127: train_loss: 2.098e-01, test_loss: 4.343e-01; train_MEE: 6.455e-01, test_MEE: 8.792e-01  lr : 0.002\n",
      "Epoch 128: train_loss: 2.066e-01, test_loss: 4.312e-01; train_MEE: 6.385e-01, test_MEE: 8.711e-01  lr : 0.002\n",
      "Epoch 129: train_loss: 2.060e-01, test_loss: 4.312e-01; train_MEE: 6.359e-01, test_MEE: 8.724e-01  lr : 0.002\n",
      "Epoch 130: train_loss: 2.066e-01, test_loss: 4.266e-01; train_MEE: 6.397e-01, test_MEE: 8.703e-01  lr : 0.002\n",
      "Epoch 131: train_loss: 2.072e-01, test_loss: 4.300e-01; train_MEE: 6.416e-01, test_MEE: 8.776e-01  lr : 0.002\n",
      "Epoch 132: train_loss: 2.055e-01, test_loss: 4.279e-01; train_MEE: 6.378e-01, test_MEE: 8.717e-01  lr : 0.002\n",
      "Epoch 133: train_loss: 2.032e-01, test_loss: 4.299e-01; train_MEE: 6.346e-01, test_MEE: 8.739e-01  lr : 0.002\n",
      "Epoch 134: train_loss: 2.030e-01, test_loss: 4.229e-01; train_MEE: 6.318e-01, test_MEE: 8.606e-01  lr : 0.002\n",
      "Epoch 135: train_loss: 2.016e-01, test_loss: 4.285e-01; train_MEE: 6.284e-01, test_MEE: 8.674e-01  lr : 0.002\n",
      "Epoch 136: train_loss: 2.054e-01, test_loss: 4.298e-01; train_MEE: 6.393e-01, test_MEE: 8.766e-01  lr : 0.002\n",
      "Epoch 137: train_loss: 2.053e-01, test_loss: 4.289e-01; train_MEE: 6.409e-01, test_MEE: 8.723e-01  lr : 0.002\n",
      "Epoch 138: train_loss: 1.999e-01, test_loss: 4.238e-01; train_MEE: 6.262e-01, test_MEE: 8.623e-01  lr : 0.002\n",
      "Epoch 139: train_loss: 2.012e-01, test_loss: 4.197e-01; train_MEE: 6.309e-01, test_MEE: 8.589e-01  lr : 0.002\n",
      "Epoch 140: train_loss: 1.988e-01, test_loss: 4.189e-01; train_MEE: 6.243e-01, test_MEE: 8.562e-01  lr : 0.002\n",
      "Epoch 141: train_loss: 1.971e-01, test_loss: 4.173e-01; train_MEE: 6.217e-01, test_MEE: 8.528e-01  lr : 0.002\n",
      "Epoch 142: train_loss: 1.980e-01, test_loss: 4.179e-01; train_MEE: 6.256e-01, test_MEE: 8.554e-01  lr : 0.002\n",
      "Epoch 143: train_loss: 1.992e-01, test_loss: 4.162e-01; train_MEE: 6.284e-01, test_MEE: 8.574e-01  lr : 0.002\n",
      "Epoch 144: train_loss: 2.016e-01, test_loss: 4.214e-01; train_MEE: 6.362e-01, test_MEE: 8.663e-01  lr : 0.002\n",
      "Epoch 145: train_loss: 2.018e-01, test_loss: 4.258e-01; train_MEE: 6.335e-01, test_MEE: 8.687e-01  lr : 0.002\n",
      "Epoch 146: train_loss: 1.953e-01, test_loss: 4.169e-01; train_MEE: 6.189e-01, test_MEE: 8.555e-01  lr : 0.002\n",
      "Epoch 147: train_loss: 1.941e-01, test_loss: 4.114e-01; train_MEE: 6.174e-01, test_MEE: 8.486e-01  lr : 0.002\n",
      "Epoch 148: train_loss: 1.937e-01, test_loss: 4.065e-01; train_MEE: 6.156e-01, test_MEE: 8.395e-01  lr : 0.002\n",
      "Epoch 149: train_loss: 1.918e-01, test_loss: 4.058e-01; train_MEE: 6.112e-01, test_MEE: 8.417e-01  lr : 0.002\n",
      "Epoch 150: train_loss: 1.911e-01, test_loss: 4.051e-01; train_MEE: 6.095e-01, test_MEE: 8.393e-01  lr : 0.002\n",
      "Epoch 151: train_loss: 1.917e-01, test_loss: 4.040e-01; train_MEE: 6.097e-01, test_MEE: 8.402e-01  lr : 0.002\n",
      "Epoch 152: train_loss: 1.929e-01, test_loss: 4.021e-01; train_MEE: 6.168e-01, test_MEE: 8.385e-01  lr : 0.002\n",
      "Epoch 153: train_loss: 1.904e-01, test_loss: 4.025e-01; train_MEE: 6.100e-01, test_MEE: 8.365e-01  lr : 0.002\n",
      "Epoch 154: train_loss: 1.899e-01, test_loss: 4.000e-01; train_MEE: 6.092e-01, test_MEE: 8.333e-01  lr : 0.002\n",
      "Epoch 155: train_loss: 1.898e-01, test_loss: 3.998e-01; train_MEE: 6.110e-01, test_MEE: 8.360e-01  lr : 0.002\n",
      "Epoch 156: train_loss: 1.886e-01, test_loss: 3.979e-01; train_MEE: 6.059e-01, test_MEE: 8.324e-01  lr : 0.002\n",
      "Epoch 157: train_loss: 1.874e-01, test_loss: 3.969e-01; train_MEE: 6.044e-01, test_MEE: 8.314e-01  lr : 0.002\n",
      "Epoch 158: train_loss: 1.891e-01, test_loss: 3.967e-01; train_MEE: 6.084e-01, test_MEE: 8.314e-01  lr : 0.002\n",
      "Epoch 159: train_loss: 1.902e-01, test_loss: 4.003e-01; train_MEE: 6.118e-01, test_MEE: 8.363e-01  lr : 0.002\n",
      "Epoch 160: train_loss: 1.894e-01, test_loss: 3.965e-01; train_MEE: 6.134e-01, test_MEE: 8.354e-01  lr : 0.002\n",
      "Epoch 161: train_loss: 1.929e-01, test_loss: 3.994e-01; train_MEE: 6.218e-01, test_MEE: 8.420e-01  lr : 0.002\n",
      "Epoch 162: train_loss: 1.858e-01, test_loss: 3.972e-01; train_MEE: 6.009e-01, test_MEE: 8.310e-01  lr : 0.002\n",
      "Epoch 163: train_loss: 1.866e-01, test_loss: 4.002e-01; train_MEE: 6.026e-01, test_MEE: 8.334e-01  lr : 0.002\n",
      "Epoch 164: train_loss: 1.850e-01, test_loss: 3.940e-01; train_MEE: 6.010e-01, test_MEE: 8.280e-01  lr : 0.002\n",
      "Epoch 165: train_loss: 1.871e-01, test_loss: 3.925e-01; train_MEE: 6.080e-01, test_MEE: 8.321e-01  lr : 0.002\n",
      "Epoch 166: train_loss: 1.861e-01, test_loss: 3.890e-01; train_MEE: 6.036e-01, test_MEE: 8.209e-01  lr : 0.002\n",
      "Epoch 167: train_loss: 1.852e-01, test_loss: 3.903e-01; train_MEE: 6.006e-01, test_MEE: 8.263e-01  lr : 0.002\n",
      "Epoch 168: train_loss: 1.834e-01, test_loss: 3.862e-01; train_MEE: 5.969e-01, test_MEE: 8.175e-01  lr : 0.002\n",
      "Epoch 169: train_loss: 1.821e-01, test_loss: 3.849e-01; train_MEE: 5.954e-01, test_MEE: 8.163e-01  lr : 0.002\n",
      "Epoch 170: train_loss: 1.819e-01, test_loss: 3.859e-01; train_MEE: 5.940e-01, test_MEE: 8.146e-01  lr : 0.002\n",
      "Epoch 171: train_loss: 1.820e-01, test_loss: 3.819e-01; train_MEE: 5.937e-01, test_MEE: 8.131e-01  lr : 0.002\n",
      "Epoch 172: train_loss: 1.820e-01, test_loss: 3.825e-01; train_MEE: 5.958e-01, test_MEE: 8.139e-01  lr : 0.002\n",
      "Epoch 173: train_loss: 1.844e-01, test_loss: 3.837e-01; train_MEE: 6.031e-01, test_MEE: 8.187e-01  lr : 0.002\n",
      "Epoch 174: train_loss: 1.822e-01, test_loss: 3.817e-01; train_MEE: 5.990e-01, test_MEE: 8.152e-01  lr : 0.002\n",
      "Epoch 175: train_loss: 1.812e-01, test_loss: 3.813e-01; train_MEE: 5.947e-01, test_MEE: 8.133e-01  lr : 0.002\n",
      "Epoch 176: train_loss: 1.806e-01, test_loss: 3.811e-01; train_MEE: 5.945e-01, test_MEE: 8.135e-01  lr : 0.002\n",
      "Epoch 177: train_loss: 1.796e-01, test_loss: 3.773e-01; train_MEE: 5.909e-01, test_MEE: 8.050e-01  lr : 0.002\n",
      "Epoch 178: train_loss: 1.805e-01, test_loss: 3.786e-01; train_MEE: 5.920e-01, test_MEE: 8.098e-01  lr : 0.002\n",
      "Epoch 179: train_loss: 1.824e-01, test_loss: 3.797e-01; train_MEE: 5.963e-01, test_MEE: 8.102e-01  lr : 0.002\n",
      "Epoch 180: train_loss: 1.867e-01, test_loss: 3.836e-01; train_MEE: 6.119e-01, test_MEE: 8.222e-01  lr : 0.002\n",
      "Epoch 181: train_loss: 1.851e-01, test_loss: 3.834e-01; train_MEE: 6.079e-01, test_MEE: 8.222e-01  lr : 0.002\n",
      "Epoch 182: train_loss: 1.822e-01, test_loss: 3.756e-01; train_MEE: 5.975e-01, test_MEE: 8.061e-01  lr : 0.002\n",
      "Epoch 183: train_loss: 1.804e-01, test_loss: 3.766e-01; train_MEE: 5.930e-01, test_MEE: 8.094e-01  lr : 0.002\n",
      "Epoch 184: train_loss: 1.772e-01, test_loss: 3.703e-01; train_MEE: 5.878e-01, test_MEE: 7.995e-01  lr : 0.002\n",
      "Epoch 185: train_loss: 1.791e-01, test_loss: 3.758e-01; train_MEE: 5.898e-01, test_MEE: 8.066e-01  lr : 0.002\n",
      "Epoch 186: train_loss: 1.788e-01, test_loss: 3.711e-01; train_MEE: 5.895e-01, test_MEE: 8.002e-01  lr : 0.002\n",
      "Epoch 187: train_loss: 1.783e-01, test_loss: 3.700e-01; train_MEE: 5.891e-01, test_MEE: 7.995e-01  lr : 0.002\n",
      "Epoch 188: train_loss: 1.790e-01, test_loss: 3.738e-01; train_MEE: 5.925e-01, test_MEE: 8.055e-01  lr : 0.002\n",
      "Epoch 189: train_loss: 1.768e-01, test_loss: 3.713e-01; train_MEE: 5.857e-01, test_MEE: 8.003e-01  lr : 0.002\n",
      "Epoch 190: train_loss: 1.751e-01, test_loss: 3.691e-01; train_MEE: 5.838e-01, test_MEE: 7.973e-01  lr : 0.002\n",
      "Epoch 191: train_loss: 1.767e-01, test_loss: 3.657e-01; train_MEE: 5.868e-01, test_MEE: 7.929e-01  lr : 0.002\n",
      "Epoch 192: train_loss: 1.789e-01, test_loss: 3.733e-01; train_MEE: 5.919e-01, test_MEE: 8.061e-01  lr : 0.002\n",
      "Epoch 193: train_loss: 1.744e-01, test_loss: 3.686e-01; train_MEE: 5.823e-01, test_MEE: 7.950e-01  lr : 0.002\n",
      "Epoch 194: train_loss: 1.756e-01, test_loss: 3.646e-01; train_MEE: 5.841e-01, test_MEE: 7.926e-01  lr : 0.002\n",
      "Epoch 195: train_loss: 1.760e-01, test_loss: 3.651e-01; train_MEE: 5.872e-01, test_MEE: 7.940e-01  lr : 0.002\n",
      "Epoch 196: train_loss: 1.822e-01, test_loss: 3.720e-01; train_MEE: 6.041e-01, test_MEE: 8.105e-01  lr : 0.002\n",
      "Epoch 197: train_loss: 1.898e-01, test_loss: 3.786e-01; train_MEE: 6.243e-01, test_MEE: 8.249e-01  lr : 0.002\n",
      "Epoch 198: train_loss: 1.871e-01, test_loss: 3.732e-01; train_MEE: 6.155e-01, test_MEE: 8.185e-01  lr : 0.002\n",
      "Epoch 199: train_loss: 1.909e-01, test_loss: 3.812e-01; train_MEE: 6.294e-01, test_MEE: 8.360e-01  lr : 0.002\n",
      "Epoch 200: train_loss: 1.788e-01, test_loss: 3.692e-01; train_MEE: 5.954e-01, test_MEE: 8.021e-01  lr : 0.002\n",
      "Epoch 201: train_loss: 1.756e-01, test_loss: 3.684e-01; train_MEE: 5.867e-01, test_MEE: 7.968e-01  lr : 0.002\n",
      "Epoch 202: train_loss: 1.726e-01, test_loss: 3.631e-01; train_MEE: 5.774e-01, test_MEE: 7.850e-01  lr : 0.002\n",
      "Epoch 203: train_loss: 1.727e-01, test_loss: 3.621e-01; train_MEE: 5.785e-01, test_MEE: 7.852e-01  lr : 0.002\n",
      "Epoch 204: train_loss: 1.714e-01, test_loss: 3.615e-01; train_MEE: 5.758e-01, test_MEE: 7.893e-01  lr : 0.002\n",
      "Epoch 205: train_loss: 1.710e-01, test_loss: 3.653e-01; train_MEE: 5.737e-01, test_MEE: 7.921e-01  lr : 0.002\n",
      "Epoch 206: train_loss: 1.716e-01, test_loss: 3.612e-01; train_MEE: 5.741e-01, test_MEE: 7.837e-01  lr : 0.002\n",
      "Epoch 207: train_loss: 1.704e-01, test_loss: 3.594e-01; train_MEE: 5.732e-01, test_MEE: 7.852e-01  lr : 0.002\n",
      "Epoch 208: train_loss: 1.708e-01, test_loss: 3.580e-01; train_MEE: 5.756e-01, test_MEE: 7.813e-01  lr : 0.002\n",
      "Epoch 209: train_loss: 1.718e-01, test_loss: 3.568e-01; train_MEE: 5.799e-01, test_MEE: 7.846e-01  lr : 0.002\n",
      "Epoch 210: train_loss: 1.731e-01, test_loss: 3.586e-01; train_MEE: 5.818e-01, test_MEE: 7.855e-01  lr : 0.002\n",
      "Epoch 211: train_loss: 1.721e-01, test_loss: 3.599e-01; train_MEE: 5.761e-01, test_MEE: 7.847e-01  lr : 0.002\n",
      "Epoch 212: train_loss: 1.692e-01, test_loss: 3.563e-01; train_MEE: 5.710e-01, test_MEE: 7.760e-01  lr : 0.002\n",
      "Epoch 213: train_loss: 1.707e-01, test_loss: 3.538e-01; train_MEE: 5.769e-01, test_MEE: 7.791e-01  lr : 0.002\n",
      "Epoch 214: train_loss: 1.700e-01, test_loss: 3.596e-01; train_MEE: 5.738e-01, test_MEE: 7.859e-01  lr : 0.002\n",
      "Epoch 215: train_loss: 1.694e-01, test_loss: 3.570e-01; train_MEE: 5.748e-01, test_MEE: 7.803e-01  lr : 0.002\n",
      "Epoch 216: train_loss: 1.699e-01, test_loss: 3.528e-01; train_MEE: 5.711e-01, test_MEE: 7.727e-01  lr : 0.002\n",
      "Epoch 217: train_loss: 1.686e-01, test_loss: 3.526e-01; train_MEE: 5.689e-01, test_MEE: 7.721e-01  lr : 0.002\n",
      "Epoch 218: train_loss: 1.713e-01, test_loss: 3.552e-01; train_MEE: 5.792e-01, test_MEE: 7.818e-01  lr : 0.002\n",
      "Epoch 219: train_loss: 1.680e-01, test_loss: 3.520e-01; train_MEE: 5.679e-01, test_MEE: 7.724e-01  lr : 0.002\n",
      "Epoch 220: train_loss: 1.664e-01, test_loss: 3.510e-01; train_MEE: 5.658e-01, test_MEE: 7.700e-01  lr : 0.002\n",
      "Epoch 221: train_loss: 1.710e-01, test_loss: 3.546e-01; train_MEE: 5.820e-01, test_MEE: 7.860e-01  lr : 0.002\n",
      "Epoch 222: train_loss: 1.759e-01, test_loss: 3.562e-01; train_MEE: 5.939e-01, test_MEE: 7.869e-01  lr : 0.002\n",
      "Epoch 223: train_loss: 1.738e-01, test_loss: 3.535e-01; train_MEE: 5.893e-01, test_MEE: 7.882e-01  lr : 0.002\n",
      "Epoch 224: train_loss: 1.680e-01, test_loss: 3.465e-01; train_MEE: 5.705e-01, test_MEE: 7.720e-01  lr : 0.002\n",
      "Epoch 225: train_loss: 1.668e-01, test_loss: 3.480e-01; train_MEE: 5.673e-01, test_MEE: 7.702e-01  lr : 0.002\n",
      "Epoch 226: train_loss: 1.678e-01, test_loss: 3.527e-01; train_MEE: 5.717e-01, test_MEE: 7.738e-01  lr : 0.002\n",
      "Epoch 227: train_loss: 1.662e-01, test_loss: 3.485e-01; train_MEE: 5.667e-01, test_MEE: 7.702e-01  lr : 0.002\n",
      "Epoch 228: train_loss: 1.657e-01, test_loss: 3.455e-01; train_MEE: 5.655e-01, test_MEE: 7.688e-01  lr : 0.002\n",
      "Epoch 229: train_loss: 1.648e-01, test_loss: 3.466e-01; train_MEE: 5.638e-01, test_MEE: 7.684e-01  lr : 0.002\n",
      "Epoch 230: train_loss: 1.652e-01, test_loss: 3.448e-01; train_MEE: 5.618e-01, test_MEE: 7.644e-01  lr : 0.002\n",
      "Epoch 231: train_loss: 1.660e-01, test_loss: 3.487e-01; train_MEE: 5.654e-01, test_MEE: 7.695e-01  lr : 0.002\n",
      "Epoch 232: train_loss: 1.666e-01, test_loss: 3.470e-01; train_MEE: 5.685e-01, test_MEE: 7.689e-01  lr : 0.002\n",
      "Epoch 233: train_loss: 1.681e-01, test_loss: 3.510e-01; train_MEE: 5.743e-01, test_MEE: 7.800e-01  lr : 0.002\n",
      "Epoch 234: train_loss: 1.747e-01, test_loss: 3.480e-01; train_MEE: 5.865e-01, test_MEE: 7.787e-01  lr : 0.002\n",
      "Epoch 235: train_loss: 1.677e-01, test_loss: 3.517e-01; train_MEE: 5.730e-01, test_MEE: 7.808e-01  lr : 0.002\n",
      "Epoch 236: train_loss: 1.666e-01, test_loss: 3.456e-01; train_MEE: 5.700e-01, test_MEE: 7.706e-01  lr : 0.002\n",
      "Epoch 237: train_loss: 1.697e-01, test_loss: 3.459e-01; train_MEE: 5.776e-01, test_MEE: 7.717e-01  lr : 0.002\n",
      "Epoch 238: train_loss: 1.640e-01, test_loss: 3.409e-01; train_MEE: 5.633e-01, test_MEE: 7.634e-01  lr : 0.002\n",
      "Epoch 239: train_loss: 1.630e-01, test_loss: 3.406e-01; train_MEE: 5.591e-01, test_MEE: 7.615e-01  lr : 0.002\n",
      "Epoch 240: train_loss: 1.631e-01, test_loss: 3.449e-01; train_MEE: 5.583e-01, test_MEE: 7.671e-01  lr : 0.002\n",
      "Epoch 241: train_loss: 1.630e-01, test_loss: 3.439e-01; train_MEE: 5.600e-01, test_MEE: 7.650e-01  lr : 0.002\n",
      "Epoch 242: train_loss: 1.622e-01, test_loss: 3.434e-01; train_MEE: 5.594e-01, test_MEE: 7.642e-01  lr : 0.002\n",
      "Epoch 243: train_loss: 1.631e-01, test_loss: 3.403e-01; train_MEE: 5.600e-01, test_MEE: 7.583e-01  lr : 0.002\n",
      "Epoch 244: train_loss: 1.612e-01, test_loss: 3.410e-01; train_MEE: 5.556e-01, test_MEE: 7.580e-01  lr : 0.002\n",
      "Epoch 245: train_loss: 1.626e-01, test_loss: 3.429e-01; train_MEE: 5.600e-01, test_MEE: 7.642e-01  lr : 0.002\n",
      "Epoch 246: train_loss: 1.655e-01, test_loss: 3.447e-01; train_MEE: 5.702e-01, test_MEE: 7.737e-01  lr : 0.002\n",
      "Epoch 247: train_loss: 1.630e-01, test_loss: 3.396e-01; train_MEE: 5.631e-01, test_MEE: 7.607e-01  lr : 0.002\n",
      "Epoch 248: train_loss: 1.632e-01, test_loss: 3.377e-01; train_MEE: 5.613e-01, test_MEE: 7.571e-01  lr : 0.002\n",
      "Epoch 249: train_loss: 1.633e-01, test_loss: 3.411e-01; train_MEE: 5.622e-01, test_MEE: 7.612e-01  lr : 0.002\n",
      "Epoch 250: train_loss: 1.640e-01, test_loss: 3.401e-01; train_MEE: 5.642e-01, test_MEE: 7.634e-01  lr : 0.002\n",
      "Epoch 251: train_loss: 1.648e-01, test_loss: 3.432e-01; train_MEE: 5.700e-01, test_MEE: 7.693e-01  lr : 0.002\n",
      "Epoch 252: train_loss: 1.686e-01, test_loss: 3.457e-01; train_MEE: 5.767e-01, test_MEE: 7.729e-01  lr : 0.002\n",
      "Epoch 253: train_loss: 1.669e-01, test_loss: 3.456e-01; train_MEE: 5.721e-01, test_MEE: 7.726e-01  lr : 0.002\n",
      "Epoch 254: train_loss: 1.612e-01, test_loss: 3.363e-01; train_MEE: 5.580e-01, test_MEE: 7.563e-01  lr : 0.002\n",
      "Epoch 255: train_loss: 1.607e-01, test_loss: 3.370e-01; train_MEE: 5.572e-01, test_MEE: 7.600e-01  lr : 0.002\n",
      "Epoch 256: train_loss: 1.594e-01, test_loss: 3.380e-01; train_MEE: 5.522e-01, test_MEE: 7.546e-01  lr : 0.002\n",
      "Epoch 257: train_loss: 1.616e-01, test_loss: 3.387e-01; train_MEE: 5.602e-01, test_MEE: 7.620e-01  lr : 0.002\n",
      "Epoch 258: train_loss: 1.687e-01, test_loss: 3.442e-01; train_MEE: 5.821e-01, test_MEE: 7.789e-01  lr : 0.002\n",
      "Epoch 259: train_loss: 1.699e-01, test_loss: 3.423e-01; train_MEE: 5.808e-01, test_MEE: 7.689e-01  lr : 0.002\n",
      "Epoch 260: train_loss: 1.620e-01, test_loss: 3.405e-01; train_MEE: 5.615e-01, test_MEE: 7.626e-01  lr : 0.002\n",
      "Epoch 261: train_loss: 1.615e-01, test_loss: 3.382e-01; train_MEE: 5.605e-01, test_MEE: 7.589e-01  lr : 0.002\n",
      "Epoch 262: train_loss: 1.593e-01, test_loss: 3.376e-01; train_MEE: 5.531e-01, test_MEE: 7.560e-01  lr : 0.002\n",
      "Epoch 263: train_loss: 1.579e-01, test_loss: 3.375e-01; train_MEE: 5.514e-01, test_MEE: 7.566e-01  lr : 0.002\n",
      "Epoch 264: train_loss: 1.583e-01, test_loss: 3.362e-01; train_MEE: 5.526e-01, test_MEE: 7.542e-01  lr : 0.002\n",
      "Epoch 265: train_loss: 1.580e-01, test_loss: 3.349e-01; train_MEE: 5.498e-01, test_MEE: 7.534e-01  lr : 0.002\n",
      "Epoch 266: train_loss: 1.589e-01, test_loss: 3.320e-01; train_MEE: 5.541e-01, test_MEE: 7.511e-01  lr : 0.002\n",
      "Epoch 267: train_loss: 1.576e-01, test_loss: 3.320e-01; train_MEE: 5.514e-01, test_MEE: 7.473e-01  lr : 0.002\n",
      "Epoch 268: train_loss: 1.569e-01, test_loss: 3.310e-01; train_MEE: 5.503e-01, test_MEE: 7.491e-01  lr : 0.002\n",
      "Epoch 269: train_loss: 1.576e-01, test_loss: 3.311e-01; train_MEE: 5.482e-01, test_MEE: 7.482e-01  lr : 0.002\n",
      "Epoch 270: train_loss: 1.577e-01, test_loss: 3.339e-01; train_MEE: 5.495e-01, test_MEE: 7.512e-01  lr : 0.002\n",
      "Epoch 271: train_loss: 1.576e-01, test_loss: 3.312e-01; train_MEE: 5.524e-01, test_MEE: 7.494e-01  lr : 0.002\n",
      "Epoch 272: train_loss: 1.570e-01, test_loss: 3.301e-01; train_MEE: 5.476e-01, test_MEE: 7.456e-01  lr : 0.002\n",
      "Epoch 273: train_loss: 1.608e-01, test_loss: 3.348e-01; train_MEE: 5.624e-01, test_MEE: 7.629e-01  lr : 0.002\n",
      "Epoch 274: train_loss: 1.593e-01, test_loss: 3.320e-01; train_MEE: 5.547e-01, test_MEE: 7.509e-01  lr : 0.002\n",
      "Epoch 275: train_loss: 1.593e-01, test_loss: 3.294e-01; train_MEE: 5.571e-01, test_MEE: 7.481e-01  lr : 0.002\n",
      "Epoch 276: train_loss: 1.578e-01, test_loss: 3.320e-01; train_MEE: 5.504e-01, test_MEE: 7.487e-01  lr : 0.002\n",
      "Epoch 277: train_loss: 1.557e-01, test_loss: 3.282e-01; train_MEE: 5.478e-01, test_MEE: 7.445e-01  lr : 0.002\n",
      "Epoch 278: train_loss: 1.568e-01, test_loss: 3.300e-01; train_MEE: 5.492e-01, test_MEE: 7.489e-01  lr : 0.002\n",
      "Epoch 279: train_loss: 1.571e-01, test_loss: 3.294e-01; train_MEE: 5.506e-01, test_MEE: 7.478e-01  lr : 0.002\n",
      "Epoch 280: train_loss: 1.575e-01, test_loss: 3.280e-01; train_MEE: 5.525e-01, test_MEE: 7.494e-01  lr : 0.002\n",
      "Epoch 281: train_loss: 1.591e-01, test_loss: 3.300e-01; train_MEE: 5.577e-01, test_MEE: 7.505e-01  lr : 0.002\n",
      "Epoch 282: train_loss: 1.573e-01, test_loss: 3.314e-01; train_MEE: 5.522e-01, test_MEE: 7.503e-01  lr : 0.002\n",
      "Epoch 283: train_loss: 1.550e-01, test_loss: 3.321e-01; train_MEE: 5.462e-01, test_MEE: 7.505e-01  lr : 0.002\n",
      "Epoch 284: train_loss: 1.601e-01, test_loss: 3.342e-01; train_MEE: 5.580e-01, test_MEE: 7.592e-01  lr : 0.002\n",
      "Epoch 285: train_loss: 1.668e-01, test_loss: 3.404e-01; train_MEE: 5.818e-01, test_MEE: 7.729e-01  lr : 0.002\n",
      "Epoch 286: train_loss: 1.619e-01, test_loss: 3.303e-01; train_MEE: 5.622e-01, test_MEE: 7.511e-01  lr : 0.002\n",
      "Epoch 287: train_loss: 1.576e-01, test_loss: 3.292e-01; train_MEE: 5.550e-01, test_MEE: 7.490e-01  lr : 0.002\n",
      "Epoch 288: train_loss: 1.554e-01, test_loss: 3.244e-01; train_MEE: 5.464e-01, test_MEE: 7.396e-01  lr : 0.002\n",
      "Epoch 289: train_loss: 1.544e-01, test_loss: 3.267e-01; train_MEE: 5.452e-01, test_MEE: 7.435e-01  lr : 0.002\n",
      "Epoch 290: train_loss: 1.554e-01, test_loss: 3.283e-01; train_MEE: 5.457e-01, test_MEE: 7.451e-01  lr : 0.002\n",
      "Epoch 291: train_loss: 1.536e-01, test_loss: 3.283e-01; train_MEE: 5.431e-01, test_MEE: 7.434e-01  lr : 0.002\n",
      "Epoch 292: train_loss: 1.535e-01, test_loss: 3.249e-01; train_MEE: 5.426e-01, test_MEE: 7.380e-01  lr : 0.002\n",
      "Epoch 293: train_loss: 1.544e-01, test_loss: 3.298e-01; train_MEE: 5.450e-01, test_MEE: 7.468e-01  lr : 0.002\n",
      "Epoch 294: train_loss: 1.535e-01, test_loss: 3.231e-01; train_MEE: 5.434e-01, test_MEE: 7.386e-01  lr : 0.002\n",
      "Epoch 295: train_loss: 1.539e-01, test_loss: 3.241e-01; train_MEE: 5.452e-01, test_MEE: 7.402e-01  lr : 0.002\n",
      "Epoch 296: train_loss: 1.540e-01, test_loss: 3.253e-01; train_MEE: 5.437e-01, test_MEE: 7.423e-01  lr : 0.002\n",
      "Epoch 297: train_loss: 1.544e-01, test_loss: 3.254e-01; train_MEE: 5.458e-01, test_MEE: 7.404e-01  lr : 0.002\n",
      "Epoch 298: train_loss: 1.549e-01, test_loss: 3.245e-01; train_MEE: 5.472e-01, test_MEE: 7.392e-01  lr : 0.002\n",
      "Epoch 299: train_loss: 1.531e-01, test_loss: 3.232e-01; train_MEE: 5.441e-01, test_MEE: 7.387e-01  lr : 0.002\n",
      "Epoch 300: train_loss: 1.530e-01, test_loss: 3.236e-01; train_MEE: 5.428e-01, test_MEE: 7.399e-01  lr : 0.002\n",
      "Epoch 301: train_loss: 1.526e-01, test_loss: 3.221e-01; train_MEE: 5.417e-01, test_MEE: 7.366e-01  lr : 0.002\n",
      "Epoch 302: train_loss: 1.523e-01, test_loss: 3.196e-01; train_MEE: 5.418e-01, test_MEE: 7.329e-01  lr : 0.002\n",
      "Epoch 303: train_loss: 1.516e-01, test_loss: 3.206e-01; train_MEE: 5.394e-01, test_MEE: 7.365e-01  lr : 0.002\n",
      "Epoch 304: train_loss: 1.522e-01, test_loss: 3.229e-01; train_MEE: 5.409e-01, test_MEE: 7.386e-01  lr : 0.002\n",
      "Epoch 305: train_loss: 1.529e-01, test_loss: 3.251e-01; train_MEE: 5.450e-01, test_MEE: 7.398e-01  lr : 0.002\n",
      "Epoch 306: train_loss: 1.543e-01, test_loss: 3.239e-01; train_MEE: 5.489e-01, test_MEE: 7.414e-01  lr : 0.002\n",
      "Epoch 307: train_loss: 1.517e-01, test_loss: 3.212e-01; train_MEE: 5.397e-01, test_MEE: 7.362e-01  lr : 0.002\n",
      "Epoch 308: train_loss: 1.522e-01, test_loss: 3.201e-01; train_MEE: 5.424e-01, test_MEE: 7.366e-01  lr : 0.002\n",
      "Epoch 309: train_loss: 1.619e-01, test_loss: 3.305e-01; train_MEE: 5.730e-01, test_MEE: 7.631e-01  lr : 0.002\n",
      "Epoch 310: train_loss: 1.702e-01, test_loss: 3.350e-01; train_MEE: 5.912e-01, test_MEE: 7.723e-01  lr : 0.002\n",
      "Epoch 311: train_loss: 1.586e-01, test_loss: 3.314e-01; train_MEE: 5.609e-01, test_MEE: 7.575e-01  lr : 0.002\n",
      "Epoch 312: train_loss: 1.550e-01, test_loss: 3.197e-01; train_MEE: 5.490e-01, test_MEE: 7.361e-01  lr : 0.002\n",
      "Epoch 313: train_loss: 1.530e-01, test_loss: 3.223e-01; train_MEE: 5.460e-01, test_MEE: 7.410e-01  lr : 0.002\n",
      "Epoch 314: train_loss: 1.512e-01, test_loss: 3.185e-01; train_MEE: 5.426e-01, test_MEE: 7.357e-01  lr : 0.002\n",
      "Epoch 315: train_loss: 1.509e-01, test_loss: 3.210e-01; train_MEE: 5.387e-01, test_MEE: 7.372e-01  lr : 0.002\n",
      "Epoch 316: train_loss: 1.507e-01, test_loss: 3.225e-01; train_MEE: 5.431e-01, test_MEE: 7.399e-01  lr : 0.002\n",
      "Epoch 317: train_loss: 1.511e-01, test_loss: 3.226e-01; train_MEE: 5.397e-01, test_MEE: 7.394e-01  lr : 0.002\n",
      "Epoch 318: train_loss: 1.504e-01, test_loss: 3.192e-01; train_MEE: 5.389e-01, test_MEE: 7.332e-01  lr : 0.002\n",
      "Epoch 319: train_loss: 1.497e-01, test_loss: 3.165e-01; train_MEE: 5.388e-01, test_MEE: 7.282e-01  lr : 0.002\n",
      "Epoch 320: train_loss: 1.499e-01, test_loss: 3.243e-01; train_MEE: 5.383e-01, test_MEE: 7.410e-01  lr : 0.002\n",
      "Epoch 321: train_loss: 1.499e-01, test_loss: 3.272e-01; train_MEE: 5.384e-01, test_MEE: 7.435e-01  lr : 0.002\n",
      "Epoch 322: train_loss: 1.519e-01, test_loss: 3.259e-01; train_MEE: 5.446e-01, test_MEE: 7.451e-01  lr : 0.002\n",
      "Epoch 323: train_loss: 1.541e-01, test_loss: 3.246e-01; train_MEE: 5.540e-01, test_MEE: 7.436e-01  lr : 0.002\n",
      "Epoch 324: train_loss: 1.507e-01, test_loss: 3.206e-01; train_MEE: 5.426e-01, test_MEE: 7.362e-01  lr : 0.002\n",
      "Epoch 325: train_loss: 1.501e-01, test_loss: 3.167e-01; train_MEE: 5.378e-01, test_MEE: 7.282e-01  lr : 0.002\n",
      "Epoch 326: train_loss: 1.504e-01, test_loss: 3.203e-01; train_MEE: 5.428e-01, test_MEE: 7.368e-01  lr : 0.002\n",
      "Epoch 327: train_loss: 1.499e-01, test_loss: 3.218e-01; train_MEE: 5.404e-01, test_MEE: 7.393e-01  lr : 0.002\n",
      "Epoch 328: train_loss: 1.485e-01, test_loss: 3.166e-01; train_MEE: 5.352e-01, test_MEE: 7.294e-01  lr : 0.002\n",
      "Epoch 329: train_loss: 1.485e-01, test_loss: 3.164e-01; train_MEE: 5.347e-01, test_MEE: 7.283e-01  lr : 0.002\n",
      "Epoch 330: train_loss: 1.484e-01, test_loss: 3.214e-01; train_MEE: 5.355e-01, test_MEE: 7.382e-01  lr : 0.002\n",
      "Epoch 331: train_loss: 1.496e-01, test_loss: 3.169e-01; train_MEE: 5.405e-01, test_MEE: 7.311e-01  lr : 0.002\n",
      "Epoch 332: train_loss: 1.515e-01, test_loss: 3.227e-01; train_MEE: 5.466e-01, test_MEE: 7.444e-01  lr : 0.002\n",
      "Epoch 333: train_loss: 1.503e-01, test_loss: 3.197e-01; train_MEE: 5.434e-01, test_MEE: 7.331e-01  lr : 0.002\n",
      "Epoch 334: train_loss: 1.484e-01, test_loss: 3.187e-01; train_MEE: 5.372e-01, test_MEE: 7.320e-01  lr : 0.002\n",
      "Epoch 335: train_loss: 1.480e-01, test_loss: 3.178e-01; train_MEE: 5.356e-01, test_MEE: 7.307e-01  lr : 0.002\n",
      "Epoch 336: train_loss: 1.495e-01, test_loss: 3.222e-01; train_MEE: 5.406e-01, test_MEE: 7.386e-01  lr : 0.002\n",
      "Epoch 337: train_loss: 1.484e-01, test_loss: 3.187e-01; train_MEE: 5.384e-01, test_MEE: 7.335e-01  lr : 0.002\n",
      "Epoch 338: train_loss: 1.483e-01, test_loss: 3.151e-01; train_MEE: 5.376e-01, test_MEE: 7.270e-01  lr : 0.002\n",
      "Epoch 339: train_loss: 1.500e-01, test_loss: 3.190e-01; train_MEE: 5.431e-01, test_MEE: 7.350e-01  lr : 0.002\n",
      "Epoch 340: train_loss: 1.480e-01, test_loss: 3.150e-01; train_MEE: 5.350e-01, test_MEE: 7.275e-01  lr : 0.002\n",
      "Epoch 341: train_loss: 1.478e-01, test_loss: 3.182e-01; train_MEE: 5.373e-01, test_MEE: 7.354e-01  lr : 0.002\n",
      "Epoch 342: train_loss: 1.474e-01, test_loss: 3.171e-01; train_MEE: 5.363e-01, test_MEE: 7.309e-01  lr : 0.002\n",
      "Epoch 343: train_loss: 1.469e-01, test_loss: 3.183e-01; train_MEE: 5.345e-01, test_MEE: 7.338e-01  lr : 0.002\n",
      "Epoch 344: train_loss: 1.494e-01, test_loss: 3.168e-01; train_MEE: 5.411e-01, test_MEE: 7.336e-01  lr : 0.002\n",
      "Epoch 345: train_loss: 1.473e-01, test_loss: 3.167e-01; train_MEE: 5.356e-01, test_MEE: 7.335e-01  lr : 0.002\n",
      "Epoch 346: train_loss: 1.480e-01, test_loss: 3.174e-01; train_MEE: 5.389e-01, test_MEE: 7.314e-01  lr : 0.002\n",
      "Epoch 347: train_loss: 1.527e-01, test_loss: 3.206e-01; train_MEE: 5.528e-01, test_MEE: 7.427e-01  lr : 0.002\n",
      "Epoch 348: train_loss: 1.474e-01, test_loss: 3.180e-01; train_MEE: 5.371e-01, test_MEE: 7.315e-01  lr : 0.002\n",
      "Epoch 349: train_loss: 1.473e-01, test_loss: 3.177e-01; train_MEE: 5.344e-01, test_MEE: 7.300e-01  lr : 0.002\n",
      "Epoch 350: train_loss: 1.469e-01, test_loss: 3.180e-01; train_MEE: 5.341e-01, test_MEE: 7.323e-01  lr : 0.002\n",
      "Epoch 351: train_loss: 1.459e-01, test_loss: 3.156e-01; train_MEE: 5.329e-01, test_MEE: 7.274e-01  lr : 0.002\n",
      "Epoch 352: train_loss: 1.462e-01, test_loss: 3.174e-01; train_MEE: 5.332e-01, test_MEE: 7.330e-01  lr : 0.002\n",
      "Epoch 353: train_loss: 1.473e-01, test_loss: 3.156e-01; train_MEE: 5.347e-01, test_MEE: 7.291e-01  lr : 0.002\n",
      "Epoch 354: train_loss: 1.463e-01, test_loss: 3.161e-01; train_MEE: 5.320e-01, test_MEE: 7.288e-01  lr : 0.002\n",
      "Epoch 355: train_loss: 1.473e-01, test_loss: 3.198e-01; train_MEE: 5.361e-01, test_MEE: 7.336e-01  lr : 0.002\n",
      "Epoch 356: train_loss: 1.474e-01, test_loss: 3.139e-01; train_MEE: 5.349e-01, test_MEE: 7.293e-01  lr : 0.002\n",
      "Epoch 357: train_loss: 1.472e-01, test_loss: 3.160e-01; train_MEE: 5.365e-01, test_MEE: 7.292e-01  lr : 0.002\n",
      "Epoch 358: train_loss: 1.457e-01, test_loss: 3.161e-01; train_MEE: 5.334e-01, test_MEE: 7.280e-01  lr : 0.002\n",
      "Epoch 359: train_loss: 1.460e-01, test_loss: 3.163e-01; train_MEE: 5.337e-01, test_MEE: 7.314e-01  lr : 0.002\n",
      "Epoch 360: train_loss: 1.454e-01, test_loss: 3.162e-01; train_MEE: 5.316e-01, test_MEE: 7.273e-01  lr : 0.002\n",
      "Epoch 361: train_loss: 1.456e-01, test_loss: 3.172e-01; train_MEE: 5.313e-01, test_MEE: 7.294e-01  lr : 0.002\n",
      "Epoch 362: train_loss: 1.454e-01, test_loss: 3.177e-01; train_MEE: 5.345e-01, test_MEE: 7.340e-01  lr : 0.002\n",
      "Epoch 363: train_loss: 1.481e-01, test_loss: 3.157e-01; train_MEE: 5.388e-01, test_MEE: 7.314e-01  lr : 0.002\n",
      "Epoch 364: train_loss: 1.483e-01, test_loss: 3.171e-01; train_MEE: 5.410e-01, test_MEE: 7.332e-01  lr : 0.002\n",
      "Epoch 365: train_loss: 1.505e-01, test_loss: 3.235e-01; train_MEE: 5.518e-01, test_MEE: 7.482e-01  lr : 0.002\n",
      "Epoch 366: train_loss: 1.525e-01, test_loss: 3.210e-01; train_MEE: 5.549e-01, test_MEE: 7.443e-01  lr : 0.002\n",
      "Epoch 367: train_loss: 1.486e-01, test_loss: 3.140e-01; train_MEE: 5.413e-01, test_MEE: 7.273e-01  lr : 0.002\n",
      "Epoch 368: train_loss: 1.462e-01, test_loss: 3.115e-01; train_MEE: 5.346e-01, test_MEE: 7.240e-01  lr : 0.002\n",
      "Epoch 369: train_loss: 1.471e-01, test_loss: 3.162e-01; train_MEE: 5.393e-01, test_MEE: 7.324e-01  lr : 0.002\n",
      "Epoch 370: train_loss: 1.455e-01, test_loss: 3.117e-01; train_MEE: 5.329e-01, test_MEE: 7.239e-01  lr : 0.002\n",
      "Epoch 371: train_loss: 1.450e-01, test_loss: 3.127e-01; train_MEE: 5.317e-01, test_MEE: 7.257e-01  lr : 0.002\n",
      "Epoch 372: train_loss: 1.447e-01, test_loss: 3.133e-01; train_MEE: 5.315e-01, test_MEE: 7.243e-01  lr : 0.002\n",
      "Epoch 373: train_loss: 1.448e-01, test_loss: 3.153e-01; train_MEE: 5.339e-01, test_MEE: 7.311e-01  lr : 0.002\n",
      "Epoch 374: train_loss: 1.450e-01, test_loss: 3.116e-01; train_MEE: 5.324e-01, test_MEE: 7.242e-01  lr : 0.002\n",
      "Epoch 375: train_loss: 1.450e-01, test_loss: 3.143e-01; train_MEE: 5.324e-01, test_MEE: 7.269e-01  lr : 0.002\n",
      "Epoch 376: train_loss: 1.451e-01, test_loss: 3.123e-01; train_MEE: 5.321e-01, test_MEE: 7.232e-01  lr : 0.002\n",
      "Epoch 377: train_loss: 1.440e-01, test_loss: 3.167e-01; train_MEE: 5.301e-01, test_MEE: 7.314e-01  lr : 0.002\n",
      "Epoch 378: train_loss: 1.459e-01, test_loss: 3.168e-01; train_MEE: 5.351e-01, test_MEE: 7.313e-01  lr : 0.002\n",
      "Epoch 379: train_loss: 1.462e-01, test_loss: 3.137e-01; train_MEE: 5.349e-01, test_MEE: 7.260e-01  lr : 0.002\n",
      "Epoch 380: train_loss: 1.444e-01, test_loss: 3.135e-01; train_MEE: 5.321e-01, test_MEE: 7.260e-01  lr : 0.002\n",
      "Epoch 381: train_loss: 1.448e-01, test_loss: 3.149e-01; train_MEE: 5.307e-01, test_MEE: 7.256e-01  lr : 0.002\n",
      "Epoch 382: train_loss: 1.440e-01, test_loss: 3.145e-01; train_MEE: 5.289e-01, test_MEE: 7.281e-01  lr : 0.002\n",
      "Epoch 383: train_loss: 1.442e-01, test_loss: 3.129e-01; train_MEE: 5.298e-01, test_MEE: 7.221e-01  lr : 0.002\n",
      "Epoch 384: train_loss: 1.444e-01, test_loss: 3.131e-01; train_MEE: 5.296e-01, test_MEE: 7.233e-01  lr : 0.002\n",
      "Epoch 385: train_loss: 1.430e-01, test_loss: 3.119e-01; train_MEE: 5.289e-01, test_MEE: 7.217e-01  lr : 0.002\n",
      "Epoch 386: train_loss: 1.438e-01, test_loss: 3.147e-01; train_MEE: 5.278e-01, test_MEE: 7.272e-01  lr : 0.002\n",
      "Epoch 387: train_loss: 1.439e-01, test_loss: 3.119e-01; train_MEE: 5.324e-01, test_MEE: 7.242e-01  lr : 0.002\n",
      "Epoch 388: train_loss: 1.429e-01, test_loss: 3.131e-01; train_MEE: 5.275e-01, test_MEE: 7.230e-01  lr : 0.002\n",
      "Epoch 389: train_loss: 1.434e-01, test_loss: 3.157e-01; train_MEE: 5.289e-01, test_MEE: 7.278e-01  lr : 0.002\n",
      "Epoch 390: train_loss: 1.429e-01, test_loss: 3.097e-01; train_MEE: 5.256e-01, test_MEE: 7.171e-01  lr : 0.002\n",
      "Epoch 391: train_loss: 1.435e-01, test_loss: 3.135e-01; train_MEE: 5.292e-01, test_MEE: 7.254e-01  lr : 0.002\n",
      "Epoch 392: train_loss: 1.433e-01, test_loss: 3.102e-01; train_MEE: 5.290e-01, test_MEE: 7.198e-01  lr : 0.002\n",
      "Epoch 393: train_loss: 1.433e-01, test_loss: 3.118e-01; train_MEE: 5.294e-01, test_MEE: 7.234e-01  lr : 0.002\n",
      "Epoch 394: train_loss: 1.439e-01, test_loss: 3.136e-01; train_MEE: 5.317e-01, test_MEE: 7.274e-01  lr : 0.002\n",
      "Epoch 395: train_loss: 1.427e-01, test_loss: 3.149e-01; train_MEE: 5.279e-01, test_MEE: 7.280e-01  lr : 0.002\n",
      "Epoch 396: train_loss: 1.426e-01, test_loss: 3.126e-01; train_MEE: 5.278e-01, test_MEE: 7.229e-01  lr : 0.002\n",
      "Epoch 397: train_loss: 1.417e-01, test_loss: 3.126e-01; train_MEE: 5.258e-01, test_MEE: 7.245e-01  lr : 0.002\n",
      "Epoch 398: train_loss: 1.441e-01, test_loss: 3.140e-01; train_MEE: 5.341e-01, test_MEE: 7.286e-01  lr : 0.002\n",
      "Epoch 399: train_loss: 1.427e-01, test_loss: 3.119e-01; train_MEE: 5.270e-01, test_MEE: 7.217e-01  lr : 0.002\n",
      "Epoch 400: train_loss: 1.423e-01, test_loss: 3.099e-01; train_MEE: 5.263e-01, test_MEE: 7.184e-01  lr : 0.002\n",
      "Epoch 401: train_loss: 1.434e-01, test_loss: 3.136e-01; train_MEE: 5.311e-01, test_MEE: 7.240e-01  lr : 0.002\n",
      "Epoch 402: train_loss: 1.451e-01, test_loss: 3.153e-01; train_MEE: 5.359e-01, test_MEE: 7.343e-01  lr : 0.002\n",
      "Epoch 403: train_loss: 1.457e-01, test_loss: 3.140e-01; train_MEE: 5.395e-01, test_MEE: 7.304e-01  lr : 0.002\n",
      "Epoch 404: train_loss: 1.473e-01, test_loss: 3.136e-01; train_MEE: 5.411e-01, test_MEE: 7.277e-01  lr : 0.002\n",
      "Epoch 405: train_loss: 1.438e-01, test_loss: 3.126e-01; train_MEE: 5.321e-01, test_MEE: 7.258e-01  lr : 0.002\n",
      "Epoch 406: train_loss: 1.422e-01, test_loss: 3.135e-01; train_MEE: 5.268e-01, test_MEE: 7.243e-01  lr : 0.002\n",
      "Epoch 407: train_loss: 1.448e-01, test_loss: 3.151e-01; train_MEE: 5.368e-01, test_MEE: 7.303e-01  lr : 0.002\n",
      "Epoch 408: train_loss: 1.512e-01, test_loss: 3.217e-01; train_MEE: 5.545e-01, test_MEE: 7.459e-01  lr : 0.002\n",
      "Epoch 409: train_loss: 1.435e-01, test_loss: 3.117e-01; train_MEE: 5.296e-01, test_MEE: 7.219e-01  lr : 0.002\n",
      "Epoch 410: train_loss: 1.421e-01, test_loss: 3.115e-01; train_MEE: 5.281e-01, test_MEE: 7.203e-01  lr : 0.002\n",
      "Epoch 411: train_loss: 1.422e-01, test_loss: 3.121e-01; train_MEE: 5.259e-01, test_MEE: 7.209e-01  lr : 0.002\n",
      "Epoch 412: train_loss: 1.416e-01, test_loss: 3.106e-01; train_MEE: 5.246e-01, test_MEE: 7.196e-01  lr : 0.002\n",
      "Epoch 413: train_loss: 1.421e-01, test_loss: 3.101e-01; train_MEE: 5.255e-01, test_MEE: 7.193e-01  lr : 0.002\n",
      "Epoch 414: train_loss: 1.417e-01, test_loss: 3.180e-01; train_MEE: 5.281e-01, test_MEE: 7.331e-01  lr : 0.002\n",
      "Epoch 415: train_loss: 1.439e-01, test_loss: 3.112e-01; train_MEE: 5.319e-01, test_MEE: 7.229e-01  lr : 0.002\n",
      "Epoch 416: train_loss: 1.450e-01, test_loss: 3.123e-01; train_MEE: 5.331e-01, test_MEE: 7.247e-01  lr : 0.002\n",
      "Epoch 417: train_loss: 1.432e-01, test_loss: 3.144e-01; train_MEE: 5.341e-01, test_MEE: 7.296e-01  lr : 0.002\n",
      "Epoch 418: train_loss: 1.424e-01, test_loss: 3.120e-01; train_MEE: 5.269e-01, test_MEE: 7.198e-01  lr : 0.002\n",
      "Epoch 419: train_loss: 1.424e-01, test_loss: 3.140e-01; train_MEE: 5.274e-01, test_MEE: 7.233e-01  lr : 0.002\n",
      "Epoch 420: train_loss: 1.424e-01, test_loss: 3.161e-01; train_MEE: 5.290e-01, test_MEE: 7.269e-01  lr : 0.002\n",
      "Epoch 421: train_loss: 1.434e-01, test_loss: 3.122e-01; train_MEE: 5.295e-01, test_MEE: 7.235e-01  lr : 0.002\n",
      "Epoch 422: train_loss: 1.422e-01, test_loss: 3.094e-01; train_MEE: 5.311e-01, test_MEE: 7.199e-01  lr : 0.002\n",
      "Epoch 423: train_loss: 1.414e-01, test_loss: 3.124e-01; train_MEE: 5.239e-01, test_MEE: 7.187e-01  lr : 0.002\n",
      "Epoch 424: train_loss: 1.427e-01, test_loss: 3.155e-01; train_MEE: 5.297e-01, test_MEE: 7.309e-01  lr : 0.002\n",
      "Epoch 425: train_loss: 1.486e-01, test_loss: 3.219e-01; train_MEE: 5.496e-01, test_MEE: 7.427e-01  lr : 0.002\n",
      "Epoch 426: train_loss: 1.460e-01, test_loss: 3.171e-01; train_MEE: 5.383e-01, test_MEE: 7.328e-01  lr : 0.002\n",
      "Epoch 427: train_loss: 1.426e-01, test_loss: 3.139e-01; train_MEE: 5.267e-01, test_MEE: 7.216e-01  lr : 0.002\n",
      "Epoch 428: train_loss: 1.408e-01, test_loss: 3.115e-01; train_MEE: 5.235e-01, test_MEE: 7.198e-01  lr : 0.002\n",
      "Epoch 429: train_loss: 1.420e-01, test_loss: 3.109e-01; train_MEE: 5.279e-01, test_MEE: 7.204e-01  lr : 0.002\n",
      "Epoch 430: train_loss: 1.459e-01, test_loss: 3.198e-01; train_MEE: 5.394e-01, test_MEE: 7.377e-01  lr : 0.002\n",
      "Epoch 431: train_loss: 1.484e-01, test_loss: 3.238e-01; train_MEE: 5.485e-01, test_MEE: 7.467e-01  lr : 0.002\n",
      "Epoch 432: train_loss: 1.437e-01, test_loss: 3.137e-01; train_MEE: 5.352e-01, test_MEE: 7.255e-01  lr : 0.002\n",
      "Epoch 433: train_loss: 1.431e-01, test_loss: 3.142e-01; train_MEE: 5.303e-01, test_MEE: 7.250e-01  lr : 0.002\n",
      "Epoch 434: train_loss: 1.428e-01, test_loss: 3.129e-01; train_MEE: 5.289e-01, test_MEE: 7.203e-01  lr : 0.002\n",
      "Epoch 435: train_loss: 1.414e-01, test_loss: 3.110e-01; train_MEE: 5.248e-01, test_MEE: 7.189e-01  lr : 0.002\n",
      "Epoch 436: train_loss: 1.402e-01, test_loss: 3.120e-01; train_MEE: 5.221e-01, test_MEE: 7.211e-01  lr : 0.002\n",
      "Epoch 437: train_loss: 1.415e-01, test_loss: 3.137e-01; train_MEE: 5.269e-01, test_MEE: 7.238e-01  lr : 0.002\n",
      "Epoch 438: train_loss: 1.435e-01, test_loss: 3.137e-01; train_MEE: 5.326e-01, test_MEE: 7.258e-01  lr : 0.002\n",
      "Epoch 439: train_loss: 1.412e-01, test_loss: 3.122e-01; train_MEE: 5.238e-01, test_MEE: 7.229e-01  lr : 0.002\n",
      "Epoch 440: train_loss: 1.426e-01, test_loss: 3.145e-01; train_MEE: 5.290e-01, test_MEE: 7.246e-01  lr : 0.002\n",
      "Epoch 441: train_loss: 1.423e-01, test_loss: 3.122e-01; train_MEE: 5.287e-01, test_MEE: 7.224e-01  lr : 0.002\n",
      "Epoch 442: train_loss: 1.415e-01, test_loss: 3.148e-01; train_MEE: 5.254e-01, test_MEE: 7.259e-01  lr : 0.002\n",
      "Epoch 443: train_loss: 1.442e-01, test_loss: 3.175e-01; train_MEE: 5.350e-01, test_MEE: 7.304e-01  lr : 0.002\n",
      "Epoch 444: train_loss: 1.419e-01, test_loss: 3.082e-01; train_MEE: 5.254e-01, test_MEE: 7.155e-01  lr : 0.002\n",
      "Epoch 445: train_loss: 1.404e-01, test_loss: 3.147e-01; train_MEE: 5.276e-01, test_MEE: 7.295e-01  lr : 0.002\n",
      "Epoch 446: train_loss: 1.432e-01, test_loss: 3.131e-01; train_MEE: 5.308e-01, test_MEE: 7.249e-01  lr : 0.002\n",
      "Epoch 447: train_loss: 1.467e-01, test_loss: 3.199e-01; train_MEE: 5.421e-01, test_MEE: 7.384e-01  lr : 0.002\n",
      "Epoch 448: train_loss: 1.474e-01, test_loss: 3.177e-01; train_MEE: 5.444e-01, test_MEE: 7.355e-01  lr : 0.002\n",
      "Epoch 449: train_loss: 1.437e-01, test_loss: 3.182e-01; train_MEE: 5.349e-01, test_MEE: 7.337e-01  lr : 0.002\n",
      "Epoch 450: train_loss: 1.411e-01, test_loss: 3.141e-01; train_MEE: 5.242e-01, test_MEE: 7.224e-01  lr : 0.002\n",
      "Epoch 451: train_loss: 1.415e-01, test_loss: 3.138e-01; train_MEE: 5.264e-01, test_MEE: 7.226e-01  lr : 0.002\n",
      "Epoch 452: train_loss: 1.406e-01, test_loss: 3.109e-01; train_MEE: 5.225e-01, test_MEE: 7.174e-01  lr : 0.002\n",
      "Epoch 453: train_loss: 1.409e-01, test_loss: 3.120e-01; train_MEE: 5.254e-01, test_MEE: 7.210e-01  lr : 0.002\n",
      "Epoch 454: train_loss: 1.434e-01, test_loss: 3.168e-01; train_MEE: 5.312e-01, test_MEE: 7.306e-01  lr : 0.002\n",
      "Epoch 455: train_loss: 1.412e-01, test_loss: 3.111e-01; train_MEE: 5.268e-01, test_MEE: 7.195e-01  lr : 0.002\n",
      "Epoch 456: train_loss: 1.411e-01, test_loss: 3.122e-01; train_MEE: 5.272e-01, test_MEE: 7.234e-01  lr : 0.002\n",
      "Epoch 457: train_loss: 1.421e-01, test_loss: 3.152e-01; train_MEE: 5.275e-01, test_MEE: 7.263e-01  lr : 0.002\n",
      "Epoch 458: train_loss: 1.411e-01, test_loss: 3.156e-01; train_MEE: 5.269e-01, test_MEE: 7.275e-01  lr : 0.002\n",
      "Epoch 459: train_loss: 1.409e-01, test_loss: 3.084e-01; train_MEE: 5.239e-01, test_MEE: 7.142e-01  lr : 0.002\n",
      "Epoch 460: train_loss: 1.395e-01, test_loss: 3.135e-01; train_MEE: 5.198e-01, test_MEE: 7.227e-01  lr : 0.002\n",
      "Epoch 461: train_loss: 1.411e-01, test_loss: 3.130e-01; train_MEE: 5.256e-01, test_MEE: 7.237e-01  lr : 0.002\n",
      "Epoch 462: train_loss: 1.426e-01, test_loss: 3.151e-01; train_MEE: 5.339e-01, test_MEE: 7.304e-01  lr : 0.002\n",
      "Epoch 463: train_loss: 1.419e-01, test_loss: 3.115e-01; train_MEE: 5.297e-01, test_MEE: 7.226e-01  lr : 0.002\n",
      "Epoch 464: train_loss: 1.413e-01, test_loss: 3.159e-01; train_MEE: 5.276e-01, test_MEE: 7.266e-01  lr : 0.002\n",
      "Epoch 465: train_loss: 1.397e-01, test_loss: 3.120e-01; train_MEE: 5.207e-01, test_MEE: 7.176e-01  lr : 0.002\n",
      "Epoch 466: train_loss: 1.387e-01, test_loss: 3.146e-01; train_MEE: 5.193e-01, test_MEE: 7.217e-01  lr : 0.002\n",
      "Epoch 467: train_loss: 1.398e-01, test_loss: 3.069e-01; train_MEE: 5.219e-01, test_MEE: 7.104e-01  lr : 0.002\n",
      "Epoch 468: train_loss: 1.389e-01, test_loss: 3.133e-01; train_MEE: 5.194e-01, test_MEE: 7.193e-01  lr : 0.002\n",
      "Epoch 469: train_loss: 1.411e-01, test_loss: 3.141e-01; train_MEE: 5.242e-01, test_MEE: 7.254e-01  lr : 0.002\n",
      "Epoch 470: train_loss: 1.406e-01, test_loss: 3.097e-01; train_MEE: 5.238e-01, test_MEE: 7.160e-01  lr : 0.002\n",
      "Epoch 471: train_loss: 1.392e-01, test_loss: 3.135e-01; train_MEE: 5.208e-01, test_MEE: 7.209e-01  lr : 0.002\n",
      "Epoch 472: train_loss: 1.389e-01, test_loss: 3.142e-01; train_MEE: 5.170e-01, test_MEE: 7.190e-01  lr : 0.002\n",
      "Epoch 473: train_loss: 1.398e-01, test_loss: 3.161e-01; train_MEE: 5.215e-01, test_MEE: 7.240e-01  lr : 0.002\n",
      "Epoch 474: train_loss: 1.398e-01, test_loss: 3.139e-01; train_MEE: 5.221e-01, test_MEE: 7.236e-01  lr : 0.002\n",
      "Epoch 475: train_loss: 1.425e-01, test_loss: 3.128e-01; train_MEE: 5.320e-01, test_MEE: 7.226e-01  lr : 0.002\n",
      "Epoch 476: train_loss: 1.401e-01, test_loss: 3.130e-01; train_MEE: 5.246e-01, test_MEE: 7.195e-01  lr : 0.002\n",
      "Epoch 477: train_loss: 1.397e-01, test_loss: 3.162e-01; train_MEE: 5.237e-01, test_MEE: 7.275e-01  lr : 0.002\n",
      "Epoch 478: train_loss: 1.398e-01, test_loss: 3.108e-01; train_MEE: 5.204e-01, test_MEE: 7.168e-01  lr : 0.002\n",
      "Epoch 479: train_loss: 1.392e-01, test_loss: 3.160e-01; train_MEE: 5.200e-01, test_MEE: 7.242e-01  lr : 0.002\n",
      "Epoch 480: train_loss: 1.389e-01, test_loss: 3.142e-01; train_MEE: 5.193e-01, test_MEE: 7.219e-01  lr : 0.002\n",
      "Epoch 481: train_loss: 1.394e-01, test_loss: 3.160e-01; train_MEE: 5.217e-01, test_MEE: 7.242e-01  lr : 0.002\n",
      "Epoch 482: train_loss: 1.395e-01, test_loss: 3.090e-01; train_MEE: 5.222e-01, test_MEE: 7.124e-01  lr : 0.002\n",
      "Epoch 483: train_loss: 1.403e-01, test_loss: 3.134e-01; train_MEE: 5.221e-01, test_MEE: 7.215e-01  lr : 0.002\n",
      "Epoch 484: train_loss: 1.421e-01, test_loss: 3.195e-01; train_MEE: 5.354e-01, test_MEE: 7.339e-01  lr : 0.002\n",
      "Epoch 485: train_loss: 1.469e-01, test_loss: 3.208e-01; train_MEE: 5.445e-01, test_MEE: 7.391e-01  lr : 0.002\n",
      "Epoch 486: train_loss: 1.493e-01, test_loss: 3.242e-01; train_MEE: 5.515e-01, test_MEE: 7.477e-01  lr : 0.002\n",
      "Epoch 487: train_loss: 1.448e-01, test_loss: 3.165e-01; train_MEE: 5.392e-01, test_MEE: 7.288e-01  lr : 0.002\n",
      "Epoch 488: train_loss: 1.410e-01, test_loss: 3.106e-01; train_MEE: 5.258e-01, test_MEE: 7.179e-01  lr : 0.002\n",
      "Epoch 489: train_loss: 1.412e-01, test_loss: 3.143e-01; train_MEE: 5.274e-01, test_MEE: 7.244e-01  lr : 0.002\n",
      "Epoch 490: train_loss: 1.392e-01, test_loss: 3.126e-01; train_MEE: 5.198e-01, test_MEE: 7.169e-01  lr : 0.002\n",
      "Epoch 491: train_loss: 1.383e-01, test_loss: 3.123e-01; train_MEE: 5.182e-01, test_MEE: 7.164e-01  lr : 0.002\n",
      "Epoch 492: train_loss: 1.404e-01, test_loss: 3.193e-01; train_MEE: 5.253e-01, test_MEE: 7.326e-01  lr : 0.002\n",
      "Epoch 493: train_loss: 1.465e-01, test_loss: 3.213e-01; train_MEE: 5.454e-01, test_MEE: 7.371e-01  lr : 0.002\n",
      "Epoch 494: train_loss: 1.440e-01, test_loss: 3.201e-01; train_MEE: 5.362e-01, test_MEE: 7.364e-01  lr : 0.002\n",
      "Epoch 495: train_loss: 1.451e-01, test_loss: 3.180e-01; train_MEE: 5.399e-01, test_MEE: 7.308e-01  lr : 0.002\n",
      "Epoch 496: train_loss: 1.430e-01, test_loss: 3.150e-01; train_MEE: 5.320e-01, test_MEE: 7.267e-01  lr : 0.002\n",
      "Epoch 497: train_loss: 1.387e-01, test_loss: 3.112e-01; train_MEE: 5.194e-01, test_MEE: 7.167e-01  lr : 0.002\n",
      "Epoch 498: train_loss: 1.402e-01, test_loss: 3.152e-01; train_MEE: 5.246e-01, test_MEE: 7.260e-01  lr : 0.002\n",
      "Epoch 499: train_loss: 1.378e-01, test_loss: 3.130e-01; train_MEE: 5.166e-01, test_MEE: 7.189e-01  lr : 0.002\n"
     ]
    }
   ],
   "source": [
    "history = model.retrain(X_train, y_train.reshape((3,X_train.shape[1])), test_data = [X_test,y_test.reshape((3,X_test.shape[1]))], **params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x253ce306a10>"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(1,figsize=(15,8))\n",
    "plt.plot(history['train_loss'],label='train_loss')\n",
    "plt.plot(history['test_loss'], label='test_loss')\n",
    "plt.yscale('log')\n",
    "plt.xlabel('Epochs', size=15)\n",
    "plt.ylabel('Loss functions', size=15)\n",
    "plt.title('train_loss vs test_loss', size=18)\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.grid()\n",
    "plt.legend(fontsize=15)\n",
    "\n",
    "plt.figure(2,figsize=(15,8))\n",
    "plt.plot(history['train_MEE'],label='train_MEE')\n",
    "plt.plot(history['test_MEE'], label='test_MEE')\n",
    "plt.yscale('log')\n",
    "plt.xlabel('Epochs', size=15)\n",
    "plt.ylabel('MEE', size=15)\n",
    "plt.title('train_MEE vs test_MEE', size=18)\n",
    "plt.xticks(size=15)\n",
    "plt.yticks(size=15)\n",
    "plt.grid()\n",
    "plt.legend(fontsize=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Blind test prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = model.predict(X_blind_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(output.T)\n",
    "df.index = df.index +1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv(f\"{DIR}data/output/output.csv\")"
   ]
  }
 ],
 "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}