tutorials/4DN_sci-Hi-C_Kim et al.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4DN sci-Hi-C cell line clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Notes\n",
    "\n",
    "This tutorial uses the new API of Higashi (wrapping all functions of Higashi into the Higashi() class).\n",
    "The old API of Higashi will still be supported and maintained).\n",
    "Please check changelog for the current status of migration from the old API to the new API."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparation\n",
    "\n",
    "### Download input files\n",
    "Download the demo data from the following link (4DN-sci-Hi-C Kim et al)\n",
    "https://drive.google.com/drive/u/1/folders/1j7ffz96kv_Ft3hicu2DRBmfcje0RA1sc\n",
    "\n",
    "The data is the same as the mtx files from https://noble.gs.washington.edu/proj/schic-topic-model/\n",
    "The mtx files are concatenated and transformed into the Higashi input format.\n",
    "\n",
    "Change the file path in the corresponding JSON file according to the location of the downloaded files.\n",
    "\n",
    "### Install Higashi\n",
    "\n",
    "1. install pytorch>=1.8.0 with cuda support when available.\n",
    "2. `conda install -c ruochiz higashi`\n",
    "(Although higashi would install pytorch when needed, there is no guarantee that it will install the correct version with cuda support. It is recommended to install pytorch separately before higashi.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Start running Higashi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Import package, set the path to the configuration JSON file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from higashi.Higashi_wrapper import *\n",
    "# Set the path to the configuration file, change it accordingly\n",
    "config = \"../config_dir/config_4DN_pub.JSON\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Process data for higashi model & prep the higashi model for training and imputation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "setting to gpu:7\n",
      "training on data from: ['chr1', 'chr2', 'chr3', 'chr4', 'chr5', 'chr6', 'chr7', 'chr8', 'chr9', 'chr10', 'chr11', 'chr12', 'chr13', 'chr14', 'chr15', 'chr16', 'chr17', 'chr18', 'chr19', 'chr20', 'chr21', 'chr22']\n",
      "total_sparsity_cell 0.012761184803150997\n",
      "no contractive loss\n",
      "Node type num [16707   499   487   397   383   362   343   319   293   283   272   271\n",
      "   268   231   215   206   181   163   157   119   127    97   103] [16707 17206 17693 18090 18473 18835 19178 19497 19790 20073 20345 20616\n",
      " 20884 21115 21330 21536 21717 21880 22037 22156 22283 22380 22483]\n",
      "start making attribute\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a73dce90ad4e413982ed495a7c70701d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss 0.8978922963142395 loss best 0.8527839779853821 epochs 104\n",
      "\n",
      "initializing data generator\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3f0a859812b64ad58e2166756e25fb6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/22 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "initializing data generator\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ae9335a0dd004c46986f4a899866214c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/22 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize the Higashi instance\n",
    "higashi_model = Higashi(config)\n",
    "\n",
    "# Data processing (only needs to be run for once)\n",
    "higashi.process_data()\n",
    "higashi_model.prep_model()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. First stage training for embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First stage training\n",
      "[ Epoch 0 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.4409, mse:  1.3129,  acc: 84.093 %, pearson: 0.378, spearman: 0.313, elapse: 134.783 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.4081,  acc: 84.835 %,pearson: 0.499, spearman: 0.421,elapse: 0.267 s\n",
      "update_rate: 1.014705\t1.088508\tupdate_rate: 0.000000\t0.000000\tpair_ratio: 0.1\t\n",
      "[ Epoch 1 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3717, mse:  1.3468,  acc: 86.998 %, pearson: 0.569, spearman: 0.477, elapse: 98.688 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3580,  acc: 88.130 %,pearson: 0.611, spearman: 0.500,elapse: 0.278 s\n",
      "update_rate: 0.715208\t15.391309\tpair_ratio: 0.2\t\n",
      "[ Epoch 2 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3547, mse:  1.5188,  acc: 88.392 %, pearson: 0.619, spearman: 0.503, elapse: 98.162 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3528,  acc: 88.772 %,pearson: 0.644, spearman: 0.507,elapse: 0.273 s\n",
      "update_rate: 0.224767\t2.200308\tpair_ratio: 0.3\t\n",
      "[ Epoch 3 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3462, mse:  1.1899,  acc: 89.147 %, pearson: 0.650, spearman: 0.514, elapse: 97.687 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3377,  acc: 89.547 %,pearson: 0.671, spearman: 0.527,elapse: 0.263 s\n",
      "update_rate: 0.110246\t0.457779\tpair_ratio: 0.4\t\n",
      "[ Epoch 4 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3421, mse:  0.9911,  acc: 89.521 %, pearson: 0.663, spearman: 0.520, elapse: 95.539 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3455,  acc: 89.338 %,pearson: 0.639, spearman: 0.518,elapse: 0.389 s\n",
      "update_rate: 0.058654\t0.198261\tpair_ratio: 0.5\t\n",
      "[ Epoch 5 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3406, mse:  0.9741,  acc: 89.660 %, pearson: 0.667, spearman: 0.522, elapse: 95.191 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3443,  acc: 89.436 %,pearson: 0.658, spearman: 0.516,elapse: 0.270 s\n",
      "update_rate: 0.029441\t0.180600\t\n",
      "[ Epoch 6 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3401, mse:  0.9724,  acc: 89.697 %, pearson: 0.671, spearman: 0.522, elapse: 106.801 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3441,  acc: 89.605 %,pearson: 0.683, spearman: 0.515,elapse: 0.279 s\n",
      "update_rate: 0.015751\t0.131197\t\n",
      "[ Epoch 7 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3384, mse:  0.9650,  acc: 89.788 %, pearson: 0.677, spearman: 0.524, elapse: 99.196 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3348,  acc: 89.754 %,pearson: 0.695, spearman: 0.529,elapse: 0.285 s\n",
      "update_rate: 0.013218\t0.139048\t\n",
      "[ Epoch 8 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3378, mse:  0.9628,  acc: 89.826 %, pearson: 0.682, spearman: 0.524, elapse: 97.853 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3356,  acc: 89.882 %,pearson: 0.685, spearman: 0.529,elapse: 0.286 s\n",
      "update_rate: 0.005916\t0.084229\t\n",
      "[ Epoch 9 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3363, mse:  0.9625,  acc: 89.893 %, pearson: 0.686, spearman: 0.526, elapse: 97.602 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3394,  acc: 89.715 %,pearson: 0.695, spearman: 0.524,elapse: 0.278 s\n",
      "update_rate: 0.005900\t0.103330\t\n",
      "[ Epoch 10 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3353, mse:  0.9602,  acc: 89.942 %, pearson: 0.690, spearman: 0.527, elapse: 96.672 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3299,  acc: 90.085 %,pearson: 0.693, spearman: 0.539,elapse: 0.326 s\n",
      "update_rate: 0.004623\t0.093469\t\n",
      "[ Epoch 11 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3343, mse:  0.9592,  acc: 89.975 %, pearson: 0.695, spearman: 0.528, elapse: 95.213 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3391,  acc: 89.863 %,pearson: 0.703, spearman: 0.523,elapse: 0.289 s\n",
      "update_rate: 0.003104\t0.070421\t\n",
      "[ Epoch 12 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3351, mse:  0.9595,  acc: 89.959 %, pearson: 0.701, spearman: 0.527, elapse: 95.941 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3331,  acc: 89.915 %,pearson: 0.688, spearman: 0.530,elapse: 0.260 s\n",
      "update_rate: 0.003943\t0.095415\t\n",
      "[ Epoch 13 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3347, mse:  0.9602,  acc: 89.976 %, pearson: 0.703, spearman: 0.527, elapse: 94.947 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3362,  acc: 89.837 %,pearson: 0.696, spearman: 0.526,elapse: 0.295 s\n",
      "update_rate: 0.003393\t0.073497\t\n",
      "[ Epoch 14 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3344, mse:  0.9605,  acc: 89.987 %, pearson: 0.704, spearman: 0.527, elapse: 94.769 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3483,  acc: 89.471 %,pearson: 0.711, spearman: 0.507,elapse: 0.292 s\n",
      "update_rate: 0.002950\t0.095878\t\n",
      "[ Epoch 15 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3335, mse:  0.9571,  acc: 90.035 %, pearson: 0.708, spearman: 0.528, elapse: 95.031 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3357,  acc: 89.811 %,pearson: 0.699, spearman: 0.528,elapse: 0.281 s\n",
      "update_rate: 0.002305\t0.105822\t\n",
      "[ Epoch 16 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3339, mse:  0.9578,  acc: 90.030 %, pearson: 0.710, spearman: 0.528, elapse: 95.036 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3355,  acc: 89.850 %,pearson: 0.723, spearman: 0.526,elapse: 0.260 s\n",
      "update_rate: 0.002262\t0.053522\t\n",
      "[ Epoch 17 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3333, mse:  0.9581,  acc: 90.036 %, pearson: 0.712, spearman: 0.528, elapse: 94.898 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3320,  acc: 89.996 %,pearson: 0.696, spearman: 0.531,elapse: 0.269 s\n",
      "update_rate: 0.002234\t0.066548\t\n",
      "[ Epoch 18 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3334, mse:  0.9580,  acc: 90.041 %, pearson: 0.709, spearman: 0.527, elapse: 94.115 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3422,  acc: 89.646 %,pearson: 0.688, spearman: 0.517,elapse: 0.262 s\n",
      "update_rate: 0.001934\t0.047479\t\n",
      "[ Epoch 19 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3324, mse:  0.9562,  acc: 90.082 %, pearson: 0.713, spearman: 0.529, elapse: 95.451 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3358,  acc: 90.067 %,pearson: 0.690, spearman: 0.524,elapse: 0.265 s\n",
      "update_rate: 0.002309\t0.079444\t\n",
      "[ Epoch 20 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3318, mse:  0.9558,  acc: 90.090 %, pearson: 0.713, spearman: 0.529, elapse: 95.824 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3340,  acc: 90.064 %,pearson: 0.698, spearman: 0.528,elapse: 0.275 s\n",
      "update_rate: 0.002342\t0.058354\t\n",
      "[ Epoch 21 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3325, mse:  0.9563,  acc: 90.070 %, pearson: 0.713, spearman: 0.528, elapse: 95.985 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3344,  acc: 89.848 %,pearson: 0.722, spearman: 0.529,elapse: 0.261 s\n",
      "update_rate: 0.001843\t0.048854\t\n",
      "[ Epoch 22 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3322, mse:  0.9552,  acc: 90.090 %, pearson: 0.718, spearman: 0.528, elapse: 95.440 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3300,  acc: 89.981 %,pearson: 0.710, spearman: 0.533,elapse: 0.258 s\n",
      "update_rate: 0.002108\t0.041794\t\n",
      "[ Epoch 23 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3306, mse:  0.9552,  acc: 90.139 %, pearson: 0.719, spearman: 0.530, elapse: 95.604 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3375,  acc: 89.904 %,pearson: 0.714, spearman: 0.524,elapse: 0.352 s\n",
      "update_rate: 0.001511\t0.043608\t\n",
      "[ Epoch 24 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3317, mse:  0.9549,  acc: 90.099 %, pearson: 0.716, spearman: 0.528, elapse: 96.319 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3326,  acc: 90.063 %,pearson: 0.776, spearman: 0.529,elapse: 0.285 s\n",
      "update_rate: 0.001422\t0.050740\t\n",
      "[ Epoch 25 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3319, mse:  0.9554,  acc: 90.094 %, pearson: 0.720, spearman: 0.528, elapse: 95.645 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3270,  acc: 90.243 %,pearson: 0.703, spearman: 0.537,elapse: 0.262 s\n",
      "update_rate: 0.001566\t0.040280\t\n",
      "[ Epoch 26 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3314, mse:  0.9547,  acc: 90.108 %, pearson: 0.719, spearman: 0.528, elapse: 94.883 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3396,  acc: 89.839 %,pearson: 0.684, spearman: 0.519,elapse: 0.253 s\n",
      "update_rate: 0.001878\t0.050545\t\n",
      "[ Epoch 27 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3321, mse:  0.9549,  acc: 90.095 %, pearson: 0.718, spearman: 0.527, elapse: 94.179 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3358,  acc: 89.798 %,pearson: 0.734, spearman: 0.523,elapse: 0.262 s\n",
      "update_rate: 0.002032\t0.053671\t\n",
      "[ Epoch 28 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3313, mse:  0.9553,  acc: 90.117 %, pearson: 0.719, spearman: 0.527, elapse: 96.451 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3259,  acc: 90.290 %,pearson: 0.740, spearman: 0.532,elapse: 0.264 s\n",
      "update_rate: 0.001503\t0.047548\t\n",
      "[ Epoch 29 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3309, mse:  0.9539,  acc: 90.122 %, pearson: 0.724, spearman: 0.527, elapse: 95.336 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3357,  acc: 89.956 %,pearson: 0.702, spearman: 0.523,elapse: 0.258 s\n",
      "update_rate: 0.001739\t0.038025\t\n",
      "[ Epoch 30 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3317, mse:  0.9540,  acc: 90.115 %, pearson: 0.721, spearman: 0.527, elapse: 95.617 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3318,  acc: 90.062 %,pearson: 0.701, spearman: 0.528,elapse: 0.258 s\n",
      "update_rate: 0.001558\t0.035019\t\n",
      "[ Epoch 31 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3306, mse:  0.9539,  acc: 90.132 %, pearson: 0.721, spearman: 0.528, elapse: 95.573 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3393,  acc: 89.936 %,pearson: 0.721, spearman: 0.521,elapse: 0.267 s\n",
      "update_rate: 0.001228\t0.036437\t\n",
      "[ Epoch 32 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3302, mse:  0.9548,  acc: 90.161 %, pearson: 0.725, spearman: 0.529, elapse: 95.482 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3348,  acc: 89.895 %,pearson: 0.704, spearman: 0.524,elapse: 0.262 s\n",
      "update_rate: 0.001097\t0.075734\t\n",
      "[ Epoch 33 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3313, mse:  0.9543,  acc: 90.109 %, pearson: 0.720, spearman: 0.527, elapse: 95.912 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3220,  acc: 90.491 %,pearson: 0.764, spearman: 0.532,elapse: 0.264 s\n",
      "update_rate: 0.001912\t0.044737\t\n",
      "[ Epoch 34 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3304, mse:  0.9533,  acc: 90.126 %, pearson: 0.720, spearman: 0.527, elapse: 95.296 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3260,  acc: 90.268 %,pearson: 0.771, spearman: 0.534,elapse: 0.262 s\n",
      "update_rate: 0.001381\t0.030251\t\n",
      "[ Epoch 35 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3308, mse:  0.9534,  acc: 90.131 %, pearson: 0.721, spearman: 0.527, elapse: 95.609 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3289,  acc: 90.159 %,pearson: 0.723, spearman: 0.532,elapse: 0.283 s\n",
      "update_rate: 0.001156\t0.040907\t\n",
      "[ Epoch 36 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3298, mse:  0.9531,  acc: 90.160 %, pearson: 0.724, spearman: 0.528, elapse: 95.597 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3346,  acc: 90.060 %,pearson: 0.709, spearman: 0.525,elapse: 0.258 s\n",
      "update_rate: 0.001236\t0.029799\t\n",
      "[ Epoch 37 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3305, mse:  0.9537,  acc: 90.128 %, pearson: 0.725, spearman: 0.527, elapse: 96.150 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3426,  acc: 89.698 %,pearson: 0.685, spearman: 0.512,elapse: 0.266 s\n",
      "update_rate: 0.001071\t0.037289\t\n",
      "[ Epoch 38 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3303, mse:  0.9551,  acc: 90.119 %, pearson: 0.723, spearman: 0.527, elapse: 95.697 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3312,  acc: 90.104 %,pearson: 0.752, spearman: 0.528,elapse: 0.266 s\n",
      "update_rate: 0.001176\t0.042577\t\n",
      "[ Epoch 39 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3307, mse:  0.9534,  acc: 90.134 %, pearson: 0.722, spearman: 0.527, elapse: 95.803 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3314,  acc: 90.082 %,pearson: 0.729, spearman: 0.525,elapse: 0.266 s\n",
      "update_rate: 0.001352\t0.039690\t\n",
      "[ Epoch 40 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3300, mse:  0.9540,  acc: 90.151 %, pearson: 0.724, spearman: 0.527, elapse: 96.547 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3343,  acc: 89.996 %,pearson: 0.695, spearman: 0.525,elapse: 0.261 s\n",
      "update_rate: 0.000931\t0.029032\t\n",
      "[ Epoch 41 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3296, mse:  0.9530,  acc: 90.128 %, pearson: 0.726, spearman: 0.527, elapse: 96.028 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3339,  acc: 89.935 %,pearson: 0.738, spearman: 0.524,elapse: 0.257 s\n",
      "update_rate: 0.001002\t0.035812\t\n",
      "[ Epoch 42 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3294, mse:  0.9534,  acc: 90.162 %, pearson: 0.724, spearman: 0.527, elapse: 96.579 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3404,  acc: 89.713 %,pearson: 0.734, spearman: 0.515,elapse: 0.266 s\n",
      "update_rate: 0.001158\t0.029915\t\n",
      "[ Epoch 43 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3294, mse:  0.9530,  acc: 90.153 %, pearson: 0.725, spearman: 0.527, elapse: 95.282 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3323,  acc: 89.940 %,pearson: 0.724, spearman: 0.524,elapse: 0.259 s\n",
      "update_rate: 0.000813\t0.022173\t\n",
      "[ Epoch 44 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3305, mse:  0.9531,  acc: 90.126 %, pearson: 0.724, spearman: 0.526, elapse: 94.964 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3383,  acc: 89.832 %,pearson: 0.694, spearman: 0.518,elapse: 0.264 s\n",
      "update_rate: 0.000904\t0.034675\t\n",
      "[ Epoch 45 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3293, mse:  0.9530,  acc: 90.186 %, pearson: 0.727, spearman: 0.528, elapse: 95.455 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3270,  acc: 90.181 %,pearson: 0.698, spearman: 0.527,elapse: 0.263 s\n",
      "update_rate: 0.001007\t0.029063\t\n",
      "[ Epoch 46 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3294, mse:  0.9529,  acc: 90.134 %, pearson: 0.726, spearman: 0.527, elapse: 94.601 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3270,  acc: 90.212 %,pearson: 0.748, spearman: 0.530,elapse: 0.263 s\n",
      "update_rate: 0.001264\t0.024716\t\n",
      "[ Epoch 47 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3293, mse:  0.9535,  acc: 90.144 %, pearson: 0.725, spearman: 0.527, elapse: 95.505 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3332,  acc: 89.834 %,pearson: 0.713, spearman: 0.519,elapse: 0.283 s\n",
      "update_rate: 0.000960\t0.032862\t\n",
      "[ Epoch 48 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3300, mse:  0.9550,  acc: 90.140 %, pearson: 0.726, spearman: 0.527, elapse: 95.268 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3343,  acc: 90.013 %,pearson: 0.709, spearman: 0.520,elapse: 0.311 s\n",
      "update_rate: 0.000971\t0.022034\t\n",
      "[ Epoch 49 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3292, mse:  0.9542,  acc: 90.140 %, pearson: 0.727, spearman: 0.526, elapse: 95.335 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3207,  acc: 90.352 %,pearson: 0.708, spearman: 0.533,elapse: 0.259 s\n",
      "update_rate: 0.000920\t0.025450\t\n",
      "[ Epoch 50 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3294, mse:  0.9534,  acc: 90.144 %, pearson: 0.724, spearman: 0.526, elapse: 95.170 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3295,  acc: 90.127 %,pearson: 0.725, spearman: 0.522,elapse: 0.285 s\n",
      "update_rate: 0.000827\t0.028155\t\n",
      "[ Epoch 51 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3288, mse:  0.9537,  acc: 90.152 %, pearson: 0.729, spearman: 0.526, elapse: 95.095 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3301,  acc: 90.094 %,pearson: 0.695, spearman: 0.525,elapse: 0.256 s\n",
      "update_rate: 0.000751\t0.026960\t\n",
      "[ Epoch 52 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3290, mse:  0.9532,  acc: 90.164 %, pearson: 0.725, spearman: 0.527, elapse: 95.888 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3240,  acc: 90.160 %,pearson: 0.707, spearman: 0.530,elapse: 0.260 s\n",
      "update_rate: 0.000972\t0.035934\t\n",
      "[ Epoch 53 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3283, mse:  0.9532,  acc: 90.168 %, pearson: 0.730, spearman: 0.528, elapse: 95.325 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3371,  acc: 89.949 %,pearson: 0.696, spearman: 0.520,elapse: 0.264 s\n",
      "update_rate: 0.000991\t0.019240\t\n",
      "[ Epoch 54 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3290, mse:  0.9539,  acc: 90.172 %, pearson: 0.726, spearman: 0.528, elapse: 95.742 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3313,  acc: 90.144 %,pearson: 0.777, spearman: 0.523,elapse: 0.259 s\n",
      "update_rate: 0.000882\t0.023413\t\n",
      "[ Epoch 55 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3281, mse:  0.9536,  acc: 90.201 %, pearson: 0.727, spearman: 0.528, elapse: 95.235 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3366,  acc: 89.887 %,pearson: 0.721, spearman: 0.521,elapse: 0.263 s\n",
      "update_rate: 0.000881\t0.030947\t\n",
      "[ Epoch 56 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3291, mse:  0.9534,  acc: 90.150 %, pearson: 0.731, spearman: 0.526, elapse: 95.181 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3298,  acc: 89.882 %,pearson: 0.698, spearman: 0.521,elapse: 0.264 s\n",
      "update_rate: 0.000933\t0.024978\t\n",
      "[ Epoch 57 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3282, mse:  0.9527,  acc: 90.183 %, pearson: 0.729, spearman: 0.528, elapse: 95.767 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3333,  acc: 89.829 %,pearson: 0.721, spearman: 0.521,elapse: 0.261 s\n",
      "update_rate: 0.001179\t0.019270\t\n",
      "[ Epoch 58 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3285, mse:  0.9617,  acc: 90.184 %, pearson: 0.730, spearman: 0.527, elapse: 95.669 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3393,  acc: 89.670 %,pearson: 0.732, spearman: 0.515,elapse: 0.259 s\n",
      "update_rate: 0.000898\t0.026274\t\n",
      "[ Epoch 59 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3291, mse:  0.9543,  acc: 90.142 %, pearson: 0.724, spearman: 0.526, elapse: 95.101 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3280,  acc: 90.018 %,pearson: 0.725, spearman: 0.524,elapse: 0.265 s\n",
      "update_rate: 0.001075\t0.022606\t\n",
      "[ Epoch 60 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3284, mse:  0.9721,  acc: 90.170 %, pearson: 0.731, spearman: 0.527, elapse: 95.249 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3335,  acc: 90.048 %,pearson: 0.735, spearman: 0.522,elapse: 0.264 s\n",
      "update_rate: 0.000799\t0.032418\t\n",
      "[ Epoch 61 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3289, mse:  0.9537,  acc: 90.143 %, pearson: 0.726, spearman: 0.526, elapse: 95.477 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3297,  acc: 90.005 %,pearson: 0.719, spearman: 0.524,elapse: 0.270 s\n",
      "update_rate: 0.000855\t0.014153\t\n",
      "[ Epoch 62 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3285, mse:  0.9549,  acc: 90.153 %, pearson: 0.728, spearman: 0.526, elapse: 95.898 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3334,  acc: 90.005 %,pearson: 0.709, spearman: 0.520,elapse: 0.286 s\n",
      "update_rate: 0.000924\t0.018809\t\n",
      "[ Epoch 63 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3273, mse:  0.9544,  acc: 90.199 %, pearson: 0.730, spearman: 0.527, elapse: 95.843 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3323,  acc: 89.913 %,pearson: 0.714, spearman: 0.522,elapse: 0.285 s\n",
      "update_rate: 0.000732\t0.017411\t\n",
      "[ Epoch 64 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3291, mse:  0.9535,  acc: 90.143 %, pearson: 0.723, spearman: 0.526, elapse: 95.758 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3404,  acc: 89.915 %,pearson: 0.700, spearman: 0.515,elapse: 0.264 s\n",
      "update_rate: 0.000978\t0.018692\t\n",
      "[ Epoch 65 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3284, mse:  0.9542,  acc: 90.173 %, pearson: 0.731, spearman: 0.526, elapse: 95.545 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3253,  acc: 90.125 %,pearson: 0.753, spearman: 0.525,elapse: 0.255 s\n",
      "update_rate: 0.000954\t0.016010\t\n",
      "[ Epoch 66 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3269, mse:  0.9537,  acc: 90.174 %, pearson: 0.731, spearman: 0.527, elapse: 95.713 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3331,  acc: 89.800 %,pearson: 0.709, spearman: 0.520,elapse: 0.257 s\n",
      "update_rate: 0.000898\t0.018034\t\n",
      "[ Epoch 67 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3291, mse:  0.9548,  acc: 90.144 %, pearson: 0.730, spearman: 0.526, elapse: 95.598 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3401,  acc: 89.846 %,pearson: 0.725, spearman: 0.515,elapse: 0.264 s\n",
      "update_rate: 0.000854\t0.017203\t\n",
      "[ Epoch 68 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3273, mse:  0.9538,  acc: 90.181 %, pearson: 0.730, spearman: 0.527, elapse: 94.893 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3276,  acc: 89.924 %,pearson: 0.718, spearman: 0.515,elapse: 0.258 s\n",
      "update_rate: 0.000967\t0.022233\t\n",
      "[ Epoch 69 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3284, mse:  0.9551,  acc: 90.148 %, pearson: 0.731, spearman: 0.526, elapse: 94.840 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3361,  acc: 89.771 %,pearson: 0.690, spearman: 0.515,elapse: 0.265 s\n",
      "update_rate: 0.000635\t0.022671\t\n",
      "[ Epoch 70 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3279, mse:  0.9577,  acc: 90.176 %, pearson: 0.730, spearman: 0.526, elapse: 95.017 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3286,  acc: 90.104 %,pearson: 0.729, spearman: 0.522,elapse: 0.264 s\n",
      "update_rate: 0.000745\t0.021528\t\n",
      "[ Epoch 71 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3282, mse:  0.9826,  acc: 90.144 %, pearson: 0.730, spearman: 0.525, elapse: 95.398 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3247,  acc: 90.200 %,pearson: 0.729, spearman: 0.527,elapse: 0.281 s\n",
      "update_rate: 0.000656\t0.020937\t\n",
      "[ Epoch 72 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3275, mse:  0.9597,  acc: 90.177 %, pearson: 0.729, spearman: 0.526, elapse: 94.920 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3358,  acc: 89.922 %,pearson: 0.730, spearman: 0.519,elapse: 0.306 s\n",
      "update_rate: 0.000799\t0.025072\t\n",
      "[ Epoch 73 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3275, mse:  0.9539,  acc: 90.169 %, pearson: 0.724, spearman: 0.525, elapse: 95.158 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3324,  acc: 90.149 %,pearson: 0.703, spearman: 0.524,elapse: 0.285 s\n",
      "update_rate: 0.000830\t0.020632\t\n",
      "[ Epoch 74 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3284, mse:  0.9544,  acc: 90.119 %, pearson: 0.729, spearman: 0.525, elapse: 94.835 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3296,  acc: 89.855 %,pearson: 0.720, spearman: 0.519,elapse: 0.258 s\n",
      "update_rate: 0.000570\t0.018636\t\n",
      "[ Epoch 75 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3272, mse:  0.9556,  acc: 90.178 %, pearson: 0.729, spearman: 0.526, elapse: 96.062 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3339,  acc: 89.947 %,pearson: 0.718, spearman: 0.522,elapse: 0.268 s\n",
      "update_rate: 0.000706\t0.017171\t\n",
      "[ Epoch 76 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3280, mse:  0.9563,  acc: 90.153 %, pearson: 0.731, spearman: 0.526, elapse: 96.144 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3297,  acc: 90.083 %,pearson: 0.727, spearman: 0.527,elapse: 0.259 s\n",
      "update_rate: 0.000698\t0.020558\t\n",
      "[ Epoch 77 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3286, mse:  0.9547,  acc: 90.130 %, pearson: 0.727, spearman: 0.525, elapse: 95.542 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3331,  acc: 90.057 %,pearson: 0.730, spearman: 0.521,elapse: 0.266 s\n",
      "update_rate: 0.000963\t0.021044\t\n",
      "[ Epoch 78 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3281, mse:  0.9561,  acc: 90.149 %, pearson: 0.731, spearman: 0.525, elapse: 94.772 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3245,  acc: 90.156 %,pearson: 0.710, spearman: 0.525,elapse: 0.256 s\n",
      "update_rate: 0.000813\t0.014690\t\n",
      "[ Epoch 79 of 80 ]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       " - (Training) :   0%|          | 0/1000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Train)   bce:  0.3273, mse:  1.0023,  acc: 90.151 %, pearson: 0.732, spearman: 0.525, elapse: 95.475 s\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- (Valid) bce:  0.3357,  acc: 90.002 %,pearson: 0.715, spearman: 0.520,elapse: 0.258 s\n",
      "update_rate: 0.000961\t0.017665\t\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  - (Validation)   :   0%|          | 0/10 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  - (Validation-hyper) bce:  0.4107,  acc: 83.215 %, auc: 0.572, aupr: 0.422,elapse: 2.350 s\n"
     ]
    }
   ],
   "source": [
    "# Stage 1 training\n",
    "higashi_model.train_for_embeddings()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Visualize embedding results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16707, 128)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize embedding results\n",
    "cell_embeddings = higashi_model.fetch_cell_embeddings()\n",
    "print (cell_embeddings.shape)\n",
    "\n",
    "from umap import UMAP\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "vec = UMAP(n_components=2, n_neighbors=25, random_state=0).fit_transform(cell_embeddings)\n",
    "cell_type = higashi_model.label_info['cell type']\n",
    "batch = higashi_model.label_info['batch']\n",
    "fig = plt.figure(figsize=(14, 5))\n",
    "ax = plt.subplot(1, 2, 1)\n",
    "sns.scatterplot(x=vec[:, 0], y=vec[:, 1], hue=cell_type, ax=ax, s=5, alpha=0.8, linewidth=0, hue_order=['GM12878','IMR90','HFF', 'HAP1', 'H1Esc'])\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "labels, handles = zip(*sorted(zip(labels, handles), key=lambda t: t[0]))\n",
    "ax.legend(handles=handles, labels=labels, bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0., ncol=1)\n",
    "ax = plt.subplot(1, 2, 2)\n",
    "sns.scatterplot(x=vec[:, 0], y=vec[:, 1], hue=batch, ax=ax, s=5, alpha=0.8, linewidth=0)\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "labels, handles = zip(*sorted(zip(labels, handles), key=lambda t: t[0]))\n",
    "ax.legend(handles=handles, labels=labels, bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0., ncol=1)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The tutorial of scHiCluster on this dataset can be found here: https://github.com/zhoujt1994/scHiCluster/tree/master/example/4DN"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}