main-LeaveOneOut.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4a4b9a3b-6422-4039-bc9c-f3ab5b9bcef2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Current device: cuda\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import numpy as np\n",
    "\n",
    "from tqdm.notebook import tqdm\n",
    "from torch.utils.tensorboard import SummaryWriter\n",
    "from torchview import draw_graph\n",
    "import graphviz\n",
    "\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "print(\"Current device:\", device)\n",
    "\n",
    "RANDOM_SEED = 1\n",
    "BATCH_SIZE = 32\n",
    "LEARN_RATE = 0.001\n",
    "MAX_NUM_EPOCH = 100\n",
    "VISUALIZE_MODEL = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b169c7fb-3232-4d05-be44-e162d542d44e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from models.CAE import CNN1D\n",
    "model = CNN1D()\n",
    "model.to(device)\n",
    "\n",
    "if VISUALIZE_MODEL:\n",
    "    model_graph = draw_graph(\n",
    "        model,\n",
    "        input_size=[(BATCH_SIZE, 84), (BATCH_SIZE, 4)],\n",
    "        expand_nested=True)\n",
    "    display(model_graph.visual_graph)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4dfce902-a39f-46d4-a30e-102ce0cb3cc3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Outcome Columns: Index(['Diagnosis', 'FEV1', 'FEV1/FVC', 'FEF2575', 'FEV1_pred',\n",
      "       'FEV1/FVC_pred', 'FEF2575_pred', 'R5', 'R5_20', 'R20', 'X5', 'AX', 'TV',\n",
      "       'R5_z_score', 'R5_20_z_score', 'R20_z_score', 'X5_z_score',\n",
      "       'AX_z_score'],\n",
      "      dtype='object')\n",
      "Info Columns: Index(['ID', 'Trial', 'Age', 'Sex', 'Height', 'Weight', 'OralCavity', 'OPJ',\n",
      "       'Glottis', 'Carina'],\n",
      "      dtype='object')\n",
      "# of total CSA samples: 1058\n",
      "# of total subjects: 340\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# of valid CSA samples w/o NaN: 524\n",
      "# of subjects w/ valid CSA samples: 209\n",
      "# of healthy samples: 295\n",
      "# of asthma sampels: 229\n"
     ]
    }
   ],
   "source": [
    "from data.aware import AwareDataset, LeaveOneOutSplitter\n",
    "\n",
    "dataset = AwareDataset(csv_data = 'data/exhale_data_v8_ave.csv',\n",
    "                       csv_outcome = 'data/exhale_outcome_v8_ave.csv',\n",
    "                       csv_info = 'data/exhale_verbose_v8_ave.csv',\n",
    "                       root_dir = 'data/',\n",
    "                       target_classes = [0,1])\n",
    "\n",
    "splitter = LeaveOneOutSplitter(dataset, BATCH_SIZE, random_seed=RANDOM_SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e56ae565-6f8f-401f-8519-5c46627859b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Fold #0\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9c6fdf920ae44e2c9e631838f2d51b39",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3f16dcead5dd4bd488199725eb13dde3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #1\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc2de7d9b6c24140aab474755667a916",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d9e4290cf94f4c7c9d033a2971ae5c44",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #2\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6da830b4418a41c98ed1c98a3881fd88",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f7ee1b12743d4e69848db826725b9c7a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #3\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "be627177cd474870990a0387eb8c77e6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "88082ac344fd4673b2a97cfccd41d938",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 20.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #4\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9d9b44205789448f9a6a8edc865ed639",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c29b20e9ee7a4f46ac80366c6d4aef52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 80.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #5\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a2a0df909ae434595dcb664c4421cea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "367a0711e7e94386b6f60658414e38c4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #6\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "40c48f0e15094f72819816f65e4f4cf4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4abe239033aa453b85e0aaa0323b3876",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #7\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b62f73868b04768880596b52728453e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fc748123ee1248a09f2e6c3e756b9c67",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #8\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c0b80188b244690ac1535eba52a3f56",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db87b2e8eb89426b826162c576a802de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #9\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "965ddb43645040919cbdc25518a39cee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "df33d5e9da80454d9ede9ff3c1f549bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #10\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a365534680fc4fc0808da4eaeed4a2b7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a4ece51a3b3b4217b03c8bddfc709d5e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #11\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c63709d1348c4f5e86851c546b230d15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c90afdb62b3b48afb74621fd35fc5fec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #12\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "89e2f46fcd384cb4ba3b653ff7e76f51",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc8a2688a97648b5915767dc7f7bad66",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #13\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e70e382ea068445fb0ea022167b7cab8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "167b174476a649f194a0cc269881b0e2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #14\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "708d9edb2be847e6a9baddd8fda9c3a3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5dfe941aa28e43298f73714fe14f82de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #15\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5211a37558dc4429b35b811d2a05a249",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "691d59099b824767915a0a9a7577e840",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #16\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ecd49499f5374440ae701b9aacec4f9e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07ce1d0a5570473295bdd0879bf5315d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #17\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7a9ed54f29fa4898a01e5cd181c92712",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a8abf05f1fe4ef3ba55dc0098a8b6da",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #18\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9da3321aaf7d408394ce305b275362c2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "de121a0663704a35b356e4b0c8d56b04",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #19\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "46d21028a9f14f3888648175b9a78f45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9638b042b7e94b82b92f96b7dc77371f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #20\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7a920a99a87b42e6b35ef4f29fae74dc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8844e55afdb941149e780c6d86174539",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #21\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "11b2731239a04ef7bc30a19a6747b361",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a0106c7032c048dc99dd6eed73dd3156",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 40.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #22\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c414bbaa79ad42cdaaaca97ee34485b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0c92065dcd984578954f6d4e294e5d17",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #23\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "acf9ce2bc6b14ab6b2d1cfd045d8c91d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1b537b368b64f8ab2bb070ef778bc8d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #24\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "93760788545a485a8b06336d9bc86e53",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "32ab5238da0b4fe4b2aee5cc52159348",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #25\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4f4d4a90f5a468da74f30e4b56baeed",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bdd028439b6f47f1bcaeb28aa1c345db",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #26\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b5afb7f39cf64fb1a19504e642de5e74",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a32a83b022f14bb59f7f92ad5c5c8bf5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #27\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "637b961236e54cdf8683da0b0358401e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "55b12fe656184a7ca8e84a8cfd815345",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #28\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2afd9e11e13641ee8961842d05bcab57",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "35ce5d7d4e3c4ba3a575b845d2425c32",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #29\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f639e251c6814251b3933b02c967588d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "143c89973e7549868da5ad3d5ce09725",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #30\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6d3db8ef2aea44d8b0982828e2c20878",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc09653298b14d118b8c75eecb34ac90",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #31\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "787a9a84dbbc4e1d91e92200d1843dd4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e95a3db6746b4125aa3ece858c36b592",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #32\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f6fd0682c8404dfea8e4d79fb713066f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "709a9bf545974d53b6af489cbd8faa53",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #33\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "001b0914874545d89481ab4a073c73b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9042dc768649473eb1dc74b34baf1648",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #34\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6716a6ba38ec45ff8b1467de2852e3d3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4786c9f393bd4e19a821ed8a32b6fba1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #35\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "86bdf3e328ec485b8e1a9af677da2718",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3903eba04b7545d1a23d721d9414638f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 75.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #36\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5eec3193dcd6475186fb0737d09a7b1d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6273b73d7f4846a5a9104281312a14f4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #37\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8160cc69c9714ed6af12f01958401fe2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a89291befd914d56991c3bf927641ed4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #38\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d678d568dd4b4a3cac1c42fdc1dab734",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fb28d43519384c6d871ad7a3bba1f237",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #39\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9560e26d8e54eb58625501143ebec45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dd56558ed1874581b92c8ccd33d79d21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #40\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e24848ec53904b43bbdadcd66f1fd58a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4ba4194180444b2693bdffb6025322f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #41\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f76c3e04e7574078833055de9797e33b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "03aed1f1cba94e74834c6d9f98633941",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #42\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a20d298cb11a45b2b1bee348627f1d75",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f56f85b98ad44e68e894ad7a7646349",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #43\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1a6eb5b2f4124eb8827d8340497e7540",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "646c6e9304de472d83492bde380e4466",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #44\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2fcd6fc1cfa0440ba2e03e488044842d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b88959b07063413299aea5dd1b914ed1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 75.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #45\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "68e75bcb5ebb4e5696eb42673ef385dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "80079160c1ae40ed9af25f6d2a952d51",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #46\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d667e3a3f76440ea8c544a5648c6f577",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aeb6eff1324146cea5887b1390908ab1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #47\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aab980eefdc242528e1f488c0f77e97e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "840172e010444fbcbe09a7e893f5170b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #48\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec2ee8628e2a425a8b27c674ddde94d9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a73b8bf311ce4c2083d47edd5de3f3f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #49\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "94d1f2c63cd043668080eeba9f97c045",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9747c87f317142348ac8dbb95a994e71",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #50\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f17cca164f184c7da8aa2fcd0632a104",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f21bea76960d4d3084fba58a152e581c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #51\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b32741ba36249cdad516a1c6762b267",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63c621da11734579964091e6919bd8e5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #52\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95b47c93b27c4784b5cf2b32a24dac09",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ae7b7ba69d46485b93e8148416f07b85",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 75.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #53\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "903b699fb1f442f1b010f703dca4eb40",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1460872b351c4934a8d5da754ee46f48",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #54\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fbf65738579c4350b70745ad7cb4144d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f2402e838dde4080bb65c4137a886fab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #55\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "35ebe42843bb435994cb9cc8e2d7acd2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63225f555a6440e399cb97413fe17914",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #56\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1f69ebcbfa2349c3853293a581cebf5a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c079ef6a34b146c2bb5492b48ce3f79e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #57\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3fc4144be448430a9d2c925643e67509",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "60b28e5a74684971a2d7b96a31215576",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #58\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8243a72917684626a2af44c38ccae3a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b5e35534b2034d728b3f4712c32e9e26",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #59\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "83d5fe7ce1a94fb88c8b2f51f755ccd6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0397da6169754301b464ce60e15ceaff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #60\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "589b4876592d4914baf13b0eb55e5a0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b7c10c07c7f42bdbf99e5e7cd91d344",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #61\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a8106b39cbfb4c569be3a70f905903fa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d7ed5ca5ab5c4e619b3b27e651e02b1e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #62\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9b353eba63a4c6d933eaeaaa7f8daa5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "236dce78491542999829902c26e9a286",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #63\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8413b668081844cfbee15e70183bb649",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4635a537104649f784f482699e95d6ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #64\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c97de0bc7e8347f293511e5956adba7e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e98b9b837e14fc5804ceb994b7ac763",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 75.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #65\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2b23a9d2c90648dca0c78cd3d19678e8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7fc1e631a8b74b42ba7c9bfcf1e92adb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #66\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9bcc30e7eee94e638da46207b137c6b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7cf90a86327d491db6e8b86b71d8ca51",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #67\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "62cc6900cbb447ccb08aeb224f872c8a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2c0b1189929141fbb83d65829b5df1c7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #68\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "604f3e16e7ba46408de1252746669f0b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "230d19361f0244d9bdbbfd9ee94b7360",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #69\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "22c2e4d7eee1460b9e67f3f859906962",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cce177c57e654df79097e5e5684c0046",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #70\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c2c4d769a12347eb8f47d31d29fe2e2e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "486f2e0ee23f40ec8065a3ccff806eac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #71\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4a8b9a0afe14ea8a454d403a5dad63b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b424a333f1934c12aa2e795f4cbe2fcc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #72\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6563c0bebc7c467ea94b3514a6de0210",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1d682ecd17974055adc594d55abd0eb1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #73\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6ce9766c9a3f4af6942099edade4e538",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d29df053dd114229ae26e3eed294075e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #74\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d3f867f2c87740c6810c8153f7ff1362",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ed59bc77fce94daea0fb9fd60ea98df6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #75\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "52a42efc545447c2a7ce802a17b15eb2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8fb5e4c4ac4b492b9d397c48755b2214",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #76\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f966cc1c9a3f49f2bffdac78b7a23c6c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d160d181b93d45e98d7e7389095fb565",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #77\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e98cbe64c48b48428a291f83023b8412",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "43384cab84bc40f4b32792da64b211c3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #78\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b43a76ddad5a4f0587168875b67d7f2c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6048e85d08434d9fa8cba34914c6e8ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #79\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b4e0d9a23b4842628bb960590cf03616",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb18d128766041a4ab1e5bb64531ade2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #80\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "723b7bc7dd424b60be91534179ecf080",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "119d5112bcdb4a6b94a8ef3c6a36571a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #81\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e685fdd423284d58ad9c3014c5e98039",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "131754f41b934ae6b8caa955b3044faf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #82\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7e0de1d980114514b53c10e7b25ef40e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "11eb5beb170842a5adfd093b9ef9cfa8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #83\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "babde3afda804e98b600c28ead8aa2dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "36ec2636581c449cb02f4d56557f18ff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #84\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bf1a5f42d04d4ba1a1be7acc2410c31b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "861bf3a6902e429eaf9bb5173766b597",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #85\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b6d3b380b11a4f9189c5052a13e84f1c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d764919258764274aa9a8f681f29e4ef",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #86\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d6a827f4fec4424aa9cf852efa4f2c20",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec8c53f628bb4e37a7a4acc3b957775b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #87\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "10166426cfde456a84ef0dea30318e98",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5939499663194606919927112910d662",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #88\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d51f2447cd374a18bcf83d424c190495",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "96db980990054e1397e673002610cf7e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #89\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8ad169d978b248938e0e7ddbeec17d2f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e8db7633b4f493f832d89c40f52d5f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 40.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #90\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ad665556f80643a5ad99dcf620964d9d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7db8c25955b04160ae7ae3d9c48b8b6e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #91\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "82097021370b436ea221d9d5a62cda12",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51f8a53a7d4a4ed3835ea12ac1d68d0d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #92\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7cee44acbaac45ef8c41644229ff04b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "45ede8463c434ccdacfcc1790d055548",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #93\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fee1300b951b4ab6aac75b981e24c36c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6d0d1b78a97e43cfa196165dbe86771a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #94\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c2c2f1b1533242d29550d484435b02f1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c7f71f22fbb54a178e21599d24bea6b1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #95\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bfcd3cf9b8914dfbb2a15729d34d1913",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0e33174f6821411ea003c0be6b6952de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #96\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e99cf7f550a94f54828fe2eff7026b32",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a914921dc80a43dfabcdc1320b8ca79d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #97\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc34d71f1e7446e4b5d1f4c31fde0622",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "279693c90db749c792a845170dc4f67b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #98\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64bd4fa12ed64fbeba1dcca33608cab6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7822c2f4abb64bc5bbf375b1f55b1e8b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #99\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8946bebfbf4e48f8a6ab40f62945b5e5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "817661c7cb44475f8cd6065e3c23e4b5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #100\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "67c23f6042254a8f8ea855892a2f41ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a8b41b11e6f3427fad6acfa68a783e74",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #101\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8373e2ac4bc442199e196d9ef6870c47",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b01967802e5442768fc1c2ee080b2a21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #102\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "481763abfc894535b2c5756bd2dd04f8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bda65775a1894dd8b808c5189de1f3e4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #103\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ef3459ae833642b8ad89e366c91f3a32",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "750d5c8379494e7bb635d069aa542c29",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #104\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc66bdc15aeb4285b50b5b256dfb68cd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3bf52ce9ac154f5e9960388cf7d060c7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #105\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1466169844e044379971c11bb3cfcd24",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1dc7237a5be4eb5a3a041fb6ff09e45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #106\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fab6c7cd8d534c27b047b2314d278a52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5e2dc028936f4b8cbd41941e67b838f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #107\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fdb9a4f27829499983d0cba132e30fa5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c0e8f4e5b7cb42fc8337d08295f3f90a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #108\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5edbe775d72a4a9bae4a6ec84cf6771c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "41118c40d2b14cac86576a71bcb46e9b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #109\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "babd1ca09e91498b96530de2a904f962",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "97fb0d5d012242bc839ed7da3af998ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 80.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #110\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ffc6943b5ce8471cb7f1ad4c1189c83c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3f59c492da284942a7756b91e328eb6a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #111\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "28543408f2fb492fa7f08d52d58f53b0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7c43325a7d3348518299024f72f5fb14",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #112\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "858686937b8e42c5ae87ff3dd2c5156c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "906804ee11374cd186630ba6a6550779",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #113\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "35a9f549c76e4718a21ecd543c709084",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ed7307acc97c4a54ba79551697920eb6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #114\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2c7b74bbd9d40a29309aa4816811934",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cbf8decb47924b3fb12901fc52208f31",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #115\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2b0e5607b4247b08c7a443392cf99dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e42ba0064d4248c387563a383fd9bc0d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #116\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb3f2b3c01c748fe9210ab604113943d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc5424b9a40142cb8e333d68632a481d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #117\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5c350d1d2c1443b7ac1efbe06f2437a4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2ccb562bbe1c459f8077c58071ca51aa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #118\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0a67e5eea0a3409cb1cbd6caa0d6d222",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "60f1fd1f75ff41a99b9ec4ac1ebdef52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #119\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8e18973d399345afa58b9bca063d7331",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "742f66cdcf2f4ebf9daab30a1724b13e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #120\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c34bb9fad4ab4e4392600da3063f6de6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3590414145aa4749ace3d5a8bc5db396",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #121\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "59f88285ebc5462ebd6db8667eab5e5f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a6648feeb6cd4531bfc65b1ab95d40a2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #122\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8362326beb49496a8aedadf5ed4f54f0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6d9be8b71f3e4ab79b6abc5ac4c2340a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 80.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #123\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "29a5c6750ffa472f925cd9993a647950",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d6a05d973d3f49e0a21677550f6ca4e7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #124\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "396725f1ab6f4917b6668105ded7df54",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4d8686e84e82408488b983ff8bb5c9ec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #125\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "98ac64d2f3d7420db4271f529bf212c1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9871734881534d0c83e77585084e3dab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #126\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9bf89c92a6244b2acba1b333ee1f0d5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2f1153c29b03469d8c31df13d386d431",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #127\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f0064c2bb03c4eb0a87c9256e8fd0253",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5bef14e47d724d1eb6d81a40badfdc19",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #128\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0bbce1d575444ac2985ac4f30c6d064e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "819eb1dea8be46a9b90071245329b5ae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #129\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "69daa606f423493384e6d7ff2209665f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ff58e381a0cb49d59d210f3145a2ac09",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #130\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc97e808aa6d4de98679e15cb5660fc7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2ee448bfd7045edaad16164b8b2e797",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #131\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5dd07225cd0f4bbc8a3b236e0b8342dc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "03587fc08cf24657a9570d6a0ef79bc6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #132\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8d3427a51832483484b03b4540df1d06",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f766cc5e8d7c49eaa6fe870e6e689503",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #133\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2ba3a71b4ded4aefba5f178633133b17",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c4370b0b7ec14007ab482ad67ab3794c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #134\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0e5250934ced4043b65af2f4c14ac802",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b1daba76ff094acb8bc05614e133c693",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #135\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "48d206936230421c8664885b79a9ac12",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e1bc587773384d44b89b52a71146ef84",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #136\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "960f562710a241899b01a186c41c8cb3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b101ee1d273246e7bf683c7e28674dbf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #137\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6da48d4d31e04acd9c08a28724822e2d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0860b51186df40d9a2602b8819468c44",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #138\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "816a5d24ff1045918b7aa43310ed7056",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e8d1c532f5bd46c29bb3336f49c44455",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #139\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a40244212dc8414fb59b6663e74a48f6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "320b83a20b944fe7ab61c8d614141155",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 40.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #140\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7605028496994a5eb282860465567585",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1fd2086e25ba4b8cb1f330308ebb083f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 75.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #141\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a511dfc7e7b84db5bf15d96fe777797c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cf7f6761422249cf8b671a6f31216d45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #142\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3ed47837f5ad45caa5ca82b2b88fc7f0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d919a1c4c2b646e2b242868a3b2252d6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #143\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e6a063860f9c41f6ab9b3336f53593d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "da5315e8a90b4cc7bb685d0b6213303b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #144\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bd5d09abcd8847b2ae6facb331e56d05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ae36f9906d3f4c5aa9cbf0f4f4a31f02",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #145\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "80a7edd3af574825bb87fce82f566130",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "259539ac4ee54c1ba2030955a91e4864",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #146\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58f62445b83247de82c48b3831d24588",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "eafdb03a15b84b50b54d2d58cd677e36",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #147\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cfe64d8df3da47139113dc5e6712255d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4e085be172494a95b6a620d0d6ff99cf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #148\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fa0d60b2b5934f40b5151a156ea51fa2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "62a914848af14d169529b892820479ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #149\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d955d1184bfe4fe49f7a41185bee00b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e90ba9c54d7e448c9d9f6f3842bdcf18",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #150\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3c10bfdea13b44aaac84773f1516b6d7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8cedc11ae0bd43e885e012dc93ba9c59",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #151\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5ae9f79793444b85a55b3fdca577b89c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a6d13413d6a47f997d0cfbd0b085a0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #152\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bee65940b5c8405e9e3893bfecbf62d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "912214143e0444b5bf533a61033e89b4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #153\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9464662ad6084ec5b251fca6749d7905",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d994b1e8c4d04acf8216b88db9efccaf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #154\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cf404b4399f84ca1987aff6969d7b766",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8e56cde74c8c493098c1fd0edefbcd04",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #155\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e1a737754dc0442bb9772ed9dd49b3ec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e9fcc275520944b3891f6c9a602c597b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #156\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1dcb3c9712c4c62b75a627b7f585555",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e32fbe4b7b6143eca68c7f19b84e8a0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #157\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3718da95b0704408bba82577b7650620",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7623f7a91fc64576b22b3b1c07768df5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #158\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d163cac70ed846e2bfadf4272af8caca",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a8cd6430c3244d58846d6863c795e59",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #159\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2913460886b04d9eb1897385f1952e91",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "21b0f79e77164e93886053c49bd8617f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #160\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1dada9c87ca54c1ca728f8eb8f46a6d8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07c6461e464b4462b93261f6351b2472",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #161\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8644b988375546a7bc3b82e9e174c553",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "378c446a67664243b93c9981337313b7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #162\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f1d1938c8bf3487c8c89db594d645521",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "be6e6f1671974cd08fdbf1256cf80929",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #163\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d56f077be0054db48fb0431afe82073a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07d71683471a46cdb2f065080b35bed1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #164\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6423cef313004c858b6a475b8f2c8d4a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "31c593e8d9f04d55b51c6108cf8a9823",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #165\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b050a43e265a46aea34d7db275f49758",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4a77f9bc3de1432fa88a226bc84b0041",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #166\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "27d1343aa81a425f892af41662fc92d5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d631d3471cac4c6d891c55672c3e04f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #167\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a7327171589448f8a4e18f1f81cf531",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "face8209560542568a9cfbdf562345bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #168\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e45c6adf1844ec49fb05580956f3cad",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0d1fcff47f17473b85740f3737c79d15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #169\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d280c91b8d5c4d61ac1ac87d7d92daa4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "35a5d7805b2e4749b62c54516420971f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #170\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "18a40aad8c0d4cef8c420cdef328e7ea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b7d6e1a4705a4dcda40d903b0f3e3157",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #171\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc82faf4791d40f58cff43fc4033ab64",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "54cdd666def44211a2d7352a9cfd5d64",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #172\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "662a618e0a13404f83ff2ccf312a636f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1df51ca0c72f48bf9d5900a238fdcd99",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #173\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "15ea4d4033ef4eb99388e2c2e90d82ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b4572099118847328feed77621a5c87a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #174\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a29f198867ab4c5d9052cf6ace112dc1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f9fe980e615841729f031bf07b414472",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #175\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "89f3cdaa1c034aabaf12541c3a82573a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "029572619c6a46b9a62202180d3386bd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #176\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1a4f6f83f404981808ab262f5f2a623",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc2bd1df7aa54a8a98197a8ea699550d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #177\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "54712e2359d94cdd83e7651087d0e9ee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "77a469c2581042b4890558d95831abe3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #178\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb8fcb27858546bca328ce8e12b18756",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58b6412eeb7644eaac6a2b7613862989",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #179\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2318a6b39ee54f4a89a5e7beb1b6ec2e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "60dc4277218c4ae79b5f7eea432acfae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #180\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ad32dcc69175411aac73bb508129e65d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "927112ed353f46998d027e9d0b145f6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #181\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2952fef74a4d44778c2f304d80e17eeb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "617e7af038564e70923dfb856bccac9b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 25.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #182\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "53929d476c034f60bf20f93481ba6080",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7a0f11e1af0540aeb1339bd4e25a8aa9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #183\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4a3eed9034584320a5cd91b41a207881",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "eef31d23cd3c4c07a54485322a603e40",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #184\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3f7af13461e94d73b288c43a9d3a782d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "66422dfe1ed04f47a4fa14fb70e103c0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #185\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ddc25681d4674844b8b377257ec9a7d9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c785bc510b2f41c7949aaad6c0151be4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #186\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7b4eabbf07e0410883d8968f0d5b227b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "751a9fedff8c41fb88fe897c09e483f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #187\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "261fff4431ae42d5b68bc6e57be1d1db",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cc3b61a95546478a90fd1023e9c112db",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #188\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07c0c540626243e1938b996710ae879b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a6652f2fe544358b3ac4913590d5bb7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #189\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "80d166586891454e813b23d5018da79f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d996fc04843945c0825e43d4b6d9f718",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #190\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "52492e06055846108e177b2873beafd7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "96e4dc0ab2304914bf40305b19bd0b0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #191\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8eec479e11e4431d8d574495d8a4ebc2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d0225d998f9542619613910fdf291566",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #192\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5172603e080046008a80582131abef06",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e45ed69f222a44ddabd5234644dab14b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #193\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "29cc11a82cd449c599bbedcd3052081f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f9a83159135f408c926e7d3b2b293af5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #194\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51d7fbeff1a8484d99e901e67c98ec06",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "384759fc88f14b16bd439190ffc86688",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #195\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6d0d97efd0e443a7a45e761a793ba7db",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "10596cf5672a4c3cae4ce05638895eab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #196\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b80c31435a6a4592925dd032b729655d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a0dd4f5233f483294ed44598e2ffad4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #197\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a0ad88ef099c40da969269332477a2af",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "123ad25bd36647f2ad725368721d5dee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #198\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1c2617b7bbd840f3862d794f64e30f27",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0a909cd79d3648e28ecd6c7d1ca1faca",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #199\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "16d4632ad6da43e590d2eb16ef57fdca",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bfd56724541b4e348b5d4a5e2b5df5f9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #200\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f50637db1f2844afbe3e748c9ad8d1f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e9fb7f1d78034f1885421acc9e803ffd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #201\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a9f8bfb9068a4e16aa4e28cfb1a887c2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9da092d6bf94e8e9720aea75f98c0eb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #202\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a16ddef416246a0b2077da6d2a1ea3a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d96e222a1d6c4d9dbd2afbf0a9a976f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #203\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7d6119ac38ba4bd3bb5a5e8db5e0875b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8cddbcfaca8e4e7ca361928aa338e86d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #204\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6047bda6fbd949ae90c49d487575e851",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8771b89f1d2f4bfead1f8525dda3eac8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #205\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "78bf1991ec7944cdbc7a69bd272b69aa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dca30cd22b5f4011a859b2f14e40d959",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #206\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "61d9357cce044b1d850ecee9e296d06c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a90b0612408490c97fa26fddf0e862a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #207\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0712f819e0c84bd6b6a83e7c86ab0bd6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "79c05d024e794350966b4c59229d748e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #208\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "41e294835cd7482eb124441e3cf5e357",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "caa62e61487e42c99f7adbe0a5a599d6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #209\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "724b14ae3ec1416da61095377286343a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f8e5567bd8eb4e23a7595cc70b446e62",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #210\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5724c5a2ff1e40f4b270358d8adffef1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "01a62749454045a4aef2000b078683ad",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #211\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3cbd8acd76484c308e4fad09bb31f346",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "03e5813c7a6c473396b3698c0314897b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #212\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "654e385485bd436a811b06f7a3313ed1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e74b4376033c4a6395e5426ee0c70248",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #213\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4a3d0c333a2e41cd8205fde7e41bcf81",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8b16945c4d5a4a44bc9b660c4f314d33",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #214\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2b6ea0a155514840992dd3ceb6c426e7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3451ff0290ea4238a0d2e3de83ca8b33",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #215\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d15ae01ce3f643a884ac8aee62bb11f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8bf56fd6939049bd82cf4a64b0c58f7c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #216\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63db32a06a8d445f8f785b6725806c64",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2bc89dbe032144c8ac60222d7e224bf9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #217\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f08f4754cc9d43eb842ab00bd9e7dcb0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "45d3248d4b7c43eb93ad726772976988",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #218\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b7ce3b25b49946be8e332a422e5f77a4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a34c148e0b444b1e9cc7dd68f46cbf6a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #219\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "040b022b2730496d9e2c93a0aa168eab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2ca9f7bb44584453810b1fdd6605bc63",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #220\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a1c9cd04b3c747f88e79687d017a5192",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "efbeda54d00e4a0fb4baf13b80fe7d8b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #221\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8acb31fa94ae440db2fdc91f9c1f7448",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "129c5d71b55b4fd0b10a2e677d3d8fc7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #222\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3a35867cf4f54823a8645d712bc47460",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e64bf5f5c63b4b94809e848ede8ef424",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #223\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58184d1c723342918c91259b2fa3d93d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b6851884523e4ef4a883577021830453",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #224\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8e975800fc3247ea97e492cfd9c7ad52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a28d03972c194a7fb527d292a03a5193",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #225\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3cee809182564be1903cdb6f66426475",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f9aeea6a77bb4fdb8b574aaec78d96be",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #226\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c2037ce006df4caea6f3ccecc2fa7d13",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b7e4472ac444423abb093dcfc6900359",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #227\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "701e1729ab3d4b2b8050e3c4d53c53f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f12b2608b7447a58667f4215a1f8f15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #228\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7f7034eb33104117bcad66d15a50d7bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3d9a32eb444f4792bbbb71aed9e4db6d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #229\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "45f5bbe9306747869517ae7da9a7db99",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b918ecff60f149498cd32b098bbc6df4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #230\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bca0af27221847caa5b74367635f8f91",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "83e723e2ae4e486f9984c75f28fca35c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #231\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "48fdf4fb40774fd5bdf7868568c6fff4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d25e598fc2fc4fd7b71cc761294d611d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #232\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e94f8c686217434bbb009af6e2b460fe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "66bcb08eb50340a0ab578396d202dc90",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #233\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a9634755d52b4ac69f015e5279c4e8a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6477d6a402df48649e3c9cdfb629f4ca",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #234\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2d24b51db5bf44729f25d3e8e15fbaba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "27e26a738dc44ccd9e383a3de0aad67d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #235\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d98555a0d3c54485ab7945513c87f36a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dd9dda3faf8242f8afd30c5e0a751d6e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #236\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "06c780df22fe491294bc416db675dff9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "34de919f1bbd4b4fad91618037091374",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #237\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e36a1aa8a3224acc8248cf9e14856bb3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d3f2cf02d8c849738c01fd4ce6f79363",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #238\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f7ebeda74122472e8ba921c010dc9297",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f00f1e95ca54535988e38d46c77e9f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #239\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e5922a48f85e4edeb41fc70cf0945dcd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bec746ba4e924a9384679a2305777e52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #240\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cfe3215ee99e4e5da057b60a6da87638",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "57e9187e72f444569ce10734a8e836cb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #241\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7236fd37d72d49f792e1c9c8b3ad06e2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c55496602aef4d56b7be7af0eb142f45",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #242\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "da6c1a7234494f10b8fd6f4740b6e59a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "924ae933b8c9409b80353cc0e50e069b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #243\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "99b134ba52784baa82cd8b875e257b6c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4f2c0a32fe204b71bcc8e57a9e12593e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 50.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #244\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dd75c788641e4a63ba4c187ead5a93cf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ee5be0e607d04a27b173c5955dda022f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #245\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cc4ad8bcbad847fa8179aa0bac0921f4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1d932d40f3e34fca8166e6c8948ba2e8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #246\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6b0522d8f9d54f2e99c004bf4c413437",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cd07fe3c14d94508bb41cadda661fd2a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #247\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b8948da4d714266884fdd1377db93bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7708603c7d204bb3b6dec56dc6f81e74",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #248\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "71c70b52a0914cedaae6093e3f031e4d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "918ec5ceb82f4aa2bd0d15dfaae8dcd6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #249\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "426dba9a082e4f2b877d84ec4cbc7161",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ae043b7ce00a4b63a1222a3b77663cb9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #250\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "710295a3408848d3ad95dd9821c60280",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f722b9834d942dabc5f351a58f3a7fb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #251\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "918999710c6b4198a4e6c049e51c6da2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f1ff523ff92147e8aa82093d95d904f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #252\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "75a95e3184ad4b5d8f481916fa1c0236",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c24ca549c9284783ab81576edd7fd305",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #253\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "47b2da200f864b7db547394dc6bd039a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bfeee14e05784e6c97cfaa3d4613b499",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #254\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db1c6f2acf174556aa6460cc31d406d8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e176ebada8564434a80b9b332c0893a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #255\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "11e4a94651c747149cf65d5c57108ad0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f19558aff58d4800af77b23e42026ad8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #256\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "50d990bec20a4ce088b80032255bb18a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7f8b44e21f0643449762d6964ee7a863",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #257\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bd1b3a2f4fb34e45826cd29de14fd4f1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "31ba6bb8209144e6a9a722b70517b7d3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 33.33%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #258\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "53e384bf9b8b42a5a76286e4a23e832a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "af143367f1674f1989cfa3c3d729f37a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #259\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4c5f09ad08246789c3a8c34fa1aa89f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ed72c3fd58ac4e9aaa96ca5d2d188695",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #260\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "322f1a236bea43b1aa101d1c78824e16",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c236812788b94c5a8ac18075979c803e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 66.67%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #261\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "878910a106c44eb39e9eff3f45f01d42",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9c4fc454b144b5895a34fa6780259dc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #262\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5284b69ae3244592ad0d5f5c21808c0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aeebf1bc55f54d0c90a44e197497d92e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #263\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a78a970577ef4f11a7d8d4fb0d60debb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c5127584b6714bd5984e7a8213d29c05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #264\n",
      "Pre-training...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0768bc84cae5410bb66b96462a00c347",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ad6ef75a1fd14cf2815311324afa09d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #265\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "26447d72450a46d2827568c868378a32",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4019c8a22b1849fab78fa2301906e88d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 0.00%\n",
      "On which epoch reach the highest accuracy: 99\n",
      "\n",
      "Fold #266\n",
      "Pre-training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ericyxy/.local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:2184: UserWarning: y_pred contains classes not in y_true\n",
      "  warnings.warn(\"y_pred contains classes not in y_true\")\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "35abb1c02d11423d980ef37ea0e18a78",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fine-tuning...\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "19227ed8723d4fda84e498c4b01ecead",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0.00/100 [00:00<?, ?epoch/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test:\n",
      "Best fine-tuning accuracy (balanced): 100.00%\n",
      "On which epoch reach the highest accuracy: 99\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "import jupyter_beeper\n",
    "\n",
    "from trainer.pretrain import PreTrainer\n",
    "from trainer.finetune import FineTuner\n",
    "from utils.others import weight_reset, Outputs\n",
    "from utils.clustering import evaluate\n",
    "from utils.outlier import novelty_detection\n",
    "\n",
    "metrics_cluster = np.zeros((5,3,3))\n",
    "outputs_test = Outputs()\n",
    "\n",
    "beeper = jupyter_beeper.Beeper()\n",
    "\n",
    "timestr = time.strftime(\"%Y%m%d-%H%M%S\")\n",
    "for split_idx, (train_loader, test_loader) in enumerate(splitter):\n",
    "    print(\"\\nFold #\" + str(split_idx))\n",
    "    writer = SummaryWriter(\"runs/\" + timestr + \"-fold\" + str(split_idx))\n",
    "    model.apply(weight_reset)\n",
    "\n",
    "    print(\"Pre-training...\")\n",
    "    model.pre_train()\n",
    "    pre_trainer = PreTrainer(\n",
    "        model,\n",
    "        lr = LEARN_RATE,\n",
    "        T_max = MAX_NUM_EPOCH,\n",
    "        device = device,\n",
    "        summarywriter = writer\n",
    "    )\n",
    "    for epoch in tqdm(range(MAX_NUM_EPOCH), unit_scale=True, unit=\"epoch\"):\n",
    "        pre_trainer.train(epoch, train_loader)\n",
    "    pre_trainer.save(pre_trainer.train_loss, epoch)\n",
    "    \n",
    "    print(\"Fine-tuning...\")\n",
    "    model.fine_tune()\n",
    "    fine_tuner = FineTuner(\n",
    "        model,\n",
    "        lr = LEARN_RATE,\n",
    "        T_max = MAX_NUM_EPOCH,\n",
    "        device = device,\n",
    "        summarywriter = writer\n",
    "    )\n",
    "    for epoch in tqdm(range(MAX_NUM_EPOCH), unit_scale=True, unit=\"epoch\"):\n",
    "        fine_tuner.train(epoch, train_loader)\n",
    "    fine_tuner.save(fine_tuner.train_acc, epoch)\n",
    "    print(\"Test:\")\n",
    "    fine_tuner.test(test_loader)\n",
    "    outputs_test.append_from(fine_tuner)\n",
    "    \n",
    "#     if split_idx == 2:\n",
    "#         beeper.beep(frequency=600, secs=0.5)\n",
    "#         break\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8e857de5-2564-48d6-9adb-0e190fe3cde4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                Labels  \\\n",
      "0    [1.0, 4.5, 80.19999694824219, 0.76999998092651...   \n",
      "1    [1.0, 2.5299999713897705, 89.4000015258789, 3....   \n",
      "2    [1.0, 3.880000114440918, 81.30000305175781, 3....   \n",
      "3    [1.0, 3.880000114440918, 81.30000305175781, 3....   \n",
      "4    [1.0, 3.880000114440918, 81.30000305175781, 3....   \n",
      "..                                                 ...   \n",
      "832  [0.0, 2.5299999713897705, 87.19999694824219, 2...   \n",
      "833  [0.0, 2.5299999713897705, 87.19999694824219, 2...   \n",
      "834  [0.0, 2.5299999713897705, 87.19999694824219, 2...   \n",
      "835  [1.0, 2.9200000762939453, 52.70000076293945, 1...   \n",
      "836  [1.0, 2.9200000762939453, 52.70000076293945, 1...   \n",
      "\n",
      "                                                  Info  \\\n",
      "0    [3.0, 1.0, 15.100000381469727, 1.0, 169.0, 87....   \n",
      "1    [4.0, 1.0, 14.699999809265137, 1.0, 157.0, 47....   \n",
      "2    [6.0, 1.0, 12.600000381469727, 0.0, 163.699996...   \n",
      "3    [6.0, 2.0, 12.600000381469727, 0.0, 163.699996...   \n",
      "4    [6.0, 3.0, 12.600000381469727, 0.0, 163.699996...   \n",
      "..                                                 ...   \n",
      "832  [365.0, 1.0, 11.800000190734863, 1.0, 151.8999...   \n",
      "833  [365.0, 2.0, 11.800000190734863, 1.0, 151.8999...   \n",
      "834  [365.0, 3.0, 11.800000190734863, 1.0, 151.8999...   \n",
      "835  [367.0, 2.0, 19.200000762939453, 1.0, 176.8000...   \n",
      "836  [367.0, 3.0, 19.200000762939453, 1.0, 176.8000...   \n",
      "\n",
      "                                Outputs_reg  \\\n",
      "0    [86.82365417480469, 85.86128997802734]   \n",
      "1    [93.69662475585938, 92.49031066894531]   \n",
      "2    [76.48789978027344, 77.96174621582031]   \n",
      "3    [75.98664855957031, 77.64790344238281]   \n",
      "4    [83.19438934326172, 84.15940856933594]   \n",
      "..                                      ...   \n",
      "832   [87.3822250366211, 86.23828887939453]   \n",
      "833    [94.2798843383789, 92.7435073852539]   \n",
      "834   [93.6355209350586, 92.11701202392578]   \n",
      "835  [92.58843994140625, 90.57131958007812]   \n",
      "836  [83.29888916015625, 82.83271789550781]   \n",
      "\n",
      "                                      Outputs_cls  \n",
      "0         [0.2149745672941208, 0.785025417804718]  \n",
      "1        [0.1026630848646164, 0.8973369002342224]  \n",
      "2     [1.1108581929875072e-06, 0.999998927116394]  \n",
      "3    [3.2380646644014632e-06, 0.9999967813491821]  \n",
      "4     [4.276277252301952e-07, 0.9999995231628418]  \n",
      "..                                            ...  \n",
      "832    [0.009297610260546207, 0.9907023906707764]  \n",
      "833    [0.002440167125314474, 0.9975597858428955]  \n",
      "834   [0.0020688087679445744, 0.9979312419891357]  \n",
      "835  [2.8989430575165898e-05, 0.9999710321426392]  \n",
      "836    [0.011795582249760628, 0.9882043600082397]  \n",
      "\n",
      "[837 rows x 4 columns]\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "print(outputs_test)\n",
    "# print(outputs_test.outputs)\n",
    "outputs_test.outputs.to_csv('data/'+timestr+'.csv', index=False)\n",
    "\n",
    "# beeper.beep(frequency=600, secs=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "id": "db2229bf-5ff6-4879-8e16-65cb2c20ed4f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     Diagnosis  Prediction  FEV1   FEV1/FVC  FEF2575  FEV1_pred  \\\n",
      "0          1.0    0.785025  4.50  80.199997     0.77      121.0   \n",
      "1          1.0    0.897337  2.53  89.400002     3.08       82.0   \n",
      "2          1.0    0.999999  3.88  81.300003     3.82      150.0   \n",
      "3          1.0    0.999997  3.88  81.300003     3.82      150.0   \n",
      "4          1.0    1.000000  3.88  81.300003     3.82      150.0   \n",
      "..         ...         ...   ...        ...      ...        ...   \n",
      "832        0.0    0.990702  2.53  87.199997     2.37       97.0   \n",
      "833        0.0    0.997560  2.53  87.199997     2.37       97.0   \n",
      "834        0.0    0.997931  2.53  87.199997     2.37       97.0   \n",
      "835        1.0    0.999971  2.92  52.700001     1.65       64.0   \n",
      "836        1.0    0.988204  2.92  52.700001     1.65       64.0   \n",
      "\n",
      "     FEV1/FVC_pred  FEF2575_pred    R5  R5_20  ...     ID  Trial        Age  \\\n",
      "0             92.0          96.0   NaN    NaN  ...    3.0    1.0  15.100000   \n",
      "1            102.0          86.0   NaN    NaN  ...    4.0    1.0  14.700000   \n",
      "2             91.0         121.0   NaN    NaN  ...    6.0    1.0  12.600000   \n",
      "3             91.0         121.0   NaN    NaN  ...    6.0    2.0  12.600000   \n",
      "4             91.0         121.0   NaN    NaN  ...    6.0    3.0  12.600000   \n",
      "..             ...           ...   ...    ...  ...    ...    ...        ...   \n",
      "832          102.0          84.0  4.50   0.47  ...  365.0    1.0  11.800000   \n",
      "833          102.0          84.0  4.50   0.47  ...  365.0    2.0  11.800000   \n",
      "834          102.0          84.0  4.50   0.47  ...  365.0    3.0  11.800000   \n",
      "835           62.0          33.0  5.65   2.27  ...  367.0    2.0  19.200001   \n",
      "836           62.0          33.0  5.65   2.27  ...  367.0    3.0  19.200001   \n",
      "\n",
      "     Sex      Height      Weight  OralCavity   OPJ  Glottis  Carina  \n",
      "0    1.0  169.000000   87.599998        46.0  55.0     91.0    95.0  \n",
      "1    1.0  157.000000   47.000000        46.0  55.0     91.0    95.0  \n",
      "2    0.0  163.699997  132.399994        47.0  66.0     88.0    99.0  \n",
      "3    0.0  163.699997  132.399994        46.0  57.0     85.0    95.0  \n",
      "4    0.0  163.699997  132.399994        49.0  67.0     88.0    95.0  \n",
      "..   ...         ...         ...         ...   ...      ...     ...  \n",
      "832  1.0  151.899994   40.700001        45.0  59.0     83.0    95.0  \n",
      "833  1.0  151.899994   40.700001        45.0  56.0     86.0    95.0  \n",
      "834  1.0  151.899994   40.700001        47.0  69.0     87.0    95.0  \n",
      "835  1.0  176.800003  128.800003        47.0  65.0     87.0   104.0  \n",
      "836  1.0  176.800003  128.800003        46.0  60.0     90.0   100.0  \n",
      "\n",
      "[837 rows x 29 columns]\n",
      "177 118\n",
      "154 388\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sens: 0.7158671586715867\n",
      "Spec: 0.6\n",
      "BalAcc: 0.6579335793357933\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import ast\n",
    "import matplotlib.pyplot as plt\n",
    "import scikitplot as skplt\n",
    "\n",
    "def convert(val):\n",
    "    return ast.literal_eval(val.replace('nan', 'None'))\n",
    "\n",
    "outputs_test.outputs = pd.read_csv(\n",
    "    'data/20231115-122738.csv', \n",
    "    converters={\n",
    "        0: convert,\n",
    "        1: convert,\n",
    "        2: convert,\n",
    "        3: convert\n",
    "    }\n",
    ")\n",
    "\n",
    "labels = outputs_test.outputs['Labels'].to_list()\n",
    "labels = pd.DataFrame(labels)\n",
    "labels.columns = ['Diagnosis', 'FEV1', 'FEV1/FVC', 'FEF2575', 'FEV1_pred',\n",
    "       'FEV1/FVC_pred', 'FEF2575_pred', 'R5', 'R5_20', 'R20', 'X5', 'AX', 'TV',\n",
    "       'R5_z_score', 'R5_20_z_score', 'R20_z_score', 'X5_z_score',\n",
    "       'AX_z_score']\n",
    "\n",
    "info = outputs_test.outputs['Info'].to_list()\n",
    "info = pd.DataFrame(info)\n",
    "info.columns = ['ID', 'Trial', 'Age', 'Sex', 'Height', 'Weight', 'OralCavity', 'OPJ',\n",
    "       'Glottis', 'Carina']\n",
    "\n",
    "meta = pd.concat([labels, info], axis=1)\n",
    "\n",
    "outputs_cls = outputs_test.outputs['Outputs_cls'].to_list()\n",
    "outputs_cls = pd.DataFrame(outputs_cls)\n",
    "\n",
    "meta.insert(1, 'Prediction', outputs_cls[1])\n",
    "\n",
    "print(meta)\n",
    "\n",
    "idx_tp = (meta['Diagnosis']==1) & (meta['Prediction']>=0.5)\n",
    "tp = idx_tp.sum()\n",
    "idx_fp = (meta['Diagnosis']==0) & (meta['Prediction']>=0.5)\n",
    "fp = idx_fp.sum()\n",
    "idx_tn = (meta['Diagnosis']==0) & (meta['Prediction']<0.5)\n",
    "tn = idx_tn.sum()\n",
    "idx_fn = (meta['Diagnosis']==1) & (meta['Prediction']<0.5)\n",
    "fn = idx_fn.sum()\n",
    "print(tn, fp)\n",
    "print(fn, tp)\n",
    "\n",
    "plt.figure()\n",
    "plt.hist(meta['Prediction'][meta['Diagnosis']==0], bins=50, alpha = 0.3, color='b', edgecolor='k', linewidth=1)\n",
    "plt.hist(meta['Prediction'][meta['Diagnosis']==1], bins=50, alpha = 0.3, color='r', edgecolor='k', linewidth=1)\n",
    "plt.legend(['Healthy', 'Asthma'])\n",
    "plt.show()\n",
    "\n",
    "# cross_entropy = -(y_true*np.log10(y_pred[1]) + (1-y_true)*np.log10(y_pred[0]))\n",
    "\n",
    "# plt.figure()\n",
    "# # plt.hist(cross_entropy, bins=np.logspace(np.log10(0.1),np.log10(10.0), 50), edgecolor='k', linewidth=1)\n",
    "# plt.hist(cross_entropy, bins=50, edgecolor='k', linewidth=1)\n",
    "# plt.legend(['Healthy', 'Asthma'])\n",
    "# # plt.gca().set_xscale(\"log\")\n",
    "# plt.vlines(-np.log10(0.5), 0, 500, color='r')\n",
    "# plt.show()\n",
    "\n",
    "# y_pred = y_pred.idxmax(axis=1)\n",
    "\n",
    "sens = tp/(tp+fn)\n",
    "spec = tn/(fp+tn)\n",
    "print('Sens:', sens)\n",
    "print('Spec:', spec)\n",
    "print('BalAcc:', (sens+spec)/2)\n",
    "\n",
    "plt.figure(figsize=(20,4))\n",
    "plt.subplot(2,4,1)\n",
    "plt.hist(pd.concat([meta['Age'][idx_tp], meta['Age'][idx_fn]], axis=1), range=(meta['Age'].min(),meta['Age'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,4,5)\n",
    "plt.hist(pd.concat([meta['Age'][idx_tn], meta['Age'][idx_fp]], axis=1), range=(meta['Age'].min(),meta['Age'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.subplot(2,4,2)\n",
    "plt.hist(pd.concat([meta['Sex'][idx_tp], meta['Sex'][idx_fn]], axis=1), range=(meta['Sex'].min(),meta['Sex'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,4,6)\n",
    "plt.hist(pd.concat([meta['Sex'][idx_tn], meta['Sex'][idx_fp]], axis=1), range=(meta['Sex'].min(),meta['Sex'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.subplot(2,4,3)\n",
    "plt.hist(pd.concat([meta['Height'][idx_tp], meta['Height'][idx_fn]], axis=1), range=(meta['Height'].min(),meta['Height'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,4,7)\n",
    "plt.hist(pd.concat([meta['Height'][idx_tn], meta['Height'][idx_fp]], axis=1), range=(meta['Height'].min(),meta['Height'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.subplot(2,4,4)\n",
    "plt.hist(pd.concat([meta['Weight'][idx_tp], meta['Weight'][idx_fn]], axis=1), range=(meta['Weight'].min(),meta['Weight'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,4,8)\n",
    "plt.hist(pd.concat([meta['Weight'][idx_tn], meta['Weight'][idx_fp]], axis=1), range=(meta['Weight'].min(),meta['Weight'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(6,4))\n",
    "plt.subplot(2,1,1)\n",
    "plt.hist(pd.concat([meta['Age'][idx_tp], meta['Age'][idx_fn]], axis=1), bins=list(range(22,40,1)), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,1,2)\n",
    "plt.hist(pd.concat([meta['Age'][idx_tn], meta['Age'][idx_fp]], axis=1), bins=list(range(22,40,1)), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(15,4))\n",
    "plt.subplot(2,3,1)\n",
    "plt.hist(pd.concat([meta['FEV1_pred'][idx_tp], meta['FEV1_pred'][idx_fn]], axis=1), range=(meta['FEV1_pred'].min(),meta['FEV1_pred'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,3,4)\n",
    "plt.hist(pd.concat([meta['FEV1_pred'][idx_tn], meta['FEV1_pred'][idx_fp]], axis=1), range=(meta['FEV1_pred'].min(),meta['FEV1_pred'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.subplot(2,3,2)\n",
    "plt.hist(pd.concat([meta['FEV1/FVC_pred'][idx_tp], meta['FEV1/FVC_pred'][idx_fn]], axis=1), range=(meta['FEV1/FVC_pred'].min(),meta['FEV1/FVC_pred'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,3,5)\n",
    "plt.hist(pd.concat([meta['FEV1/FVC_pred'][idx_tn], meta['FEV1/FVC_pred'][idx_fp]], axis=1), range=(meta['FEV1/FVC_pred'].min(),meta['FEV1/FVC_pred'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.subplot(2,3,3)\n",
    "plt.hist(pd.concat([meta['FEF2575'][idx_tp], meta['FEF2575'][idx_fn]], axis=1), range=(meta['FEF2575'].min(),meta['FEF2575'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Asthma', 'False Healthy'])\n",
    "plt.subplot(2,3,6)\n",
    "plt.hist(pd.concat([meta['FEF2575'][idx_tn], meta['FEF2575'][idx_fp]], axis=1), range=(meta['FEF2575'].min(),meta['FEF2575'].max()), alpha = 0.3, stacked=True, edgecolor='k', linewidth=1)\n",
    "plt.legend(['True Healthy', 'False Asthma'])\n",
    "plt.show()\n",
    "\n",
    "thre = np.zeros(1001)\n",
    "sens = np.zeros(1001)\n",
    "spec = np.zeros(1001)\n",
    "for i in range(0,1001):\n",
    "    idx_tp = (meta['Diagnosis']==1) & (meta['Prediction']>=i/1000) & (meta['Age']>=6) & (meta['Age']<=22)\n",
    "    tp = idx_tp.sum()\n",
    "    idx_fp = (meta['Diagnosis']==0) & (meta['Prediction']>=i/1000) & (meta['Age']>=6) & (meta['Age']<=22)\n",
    "    fp = idx_fp.sum()\n",
    "    idx_tn = (meta['Diagnosis']==0) & (meta['Prediction']<i/1000) & (meta['Age']>=6) & (meta['Age']<=22)\n",
    "    tn = idx_tn.sum()\n",
    "    idx_fn = (meta['Diagnosis']==1) & (meta['Prediction']<i/1000) & (meta['Age']>=6) & (meta['Age']<=22)\n",
    "    fn = idx_fn.sum()\n",
    "    thre[i] = i/1000\n",
    "    sens[i] = tp/(tp+fn)\n",
    "    spec[i] = tn/(fp+tn)\n",
    "    \n",
    "plt.figure()\n",
    "plt.plot(thre, sens)\n",
    "plt.plot(thre, spec)\n",
    "plt.title('Sensitivity and Specificity vs. Threshold')\n",
    "plt.legend(['Sensitivity', 'Specificity'])\n",
    "plt.xlabel('Threshold')\n",
    "plt.show()\n",
    "\n",
    "plt.figure()\n",
    "# skplt.metrics.plot_roc_curve(meta['Diagnosis'], outputs_cls)\n",
    "plt.plot(1-spec, sens)\n",
    "plt.plot([0,1], [0,1], '--k')\n",
    "plt.title('ROC Curve')\n",
    "plt.xlabel('False Positive Rate\\n(1-Specificity)')\n",
    "plt.ylabel('True Positive Rate\\n(Sensitivity)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "7bf58c48-c3f9-4f3a-a76e-dbdc0e9d9134",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 91 nearest neighbors...\n",
      "[t-SNE] Indexed 837 samples in 0.000s...\n",
      "[t-SNE] Computed neighbors for 837 samples in 0.035s...\n",
      "[t-SNE] Computed conditional probabilities for sample 837 / 837\n",
      "[t-SNE] Mean sigma: 1.924096\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 66.872063\n",
      "[t-SNE] KL divergence after 300 iterations: 1.565472\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/mnt/d/Xiangyu/AWARE/utils/clustering.py:167: UserWarning: The palette list has more values (8) than needed (2), which may not be intended.\n",
      "  sns.scatterplot(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 91 nearest neighbors...\n",
      "[t-SNE] Indexed 837 samples in 0.001s...\n",
      "[t-SNE] Computed neighbors for 837 samples in 0.029s...\n",
      "[t-SNE] Computed conditional probabilities for sample 837 / 837\n",
      "[t-SNE] Mean sigma: 1.924096\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 66.876175\n",
      "[t-SNE] KL divergence after 300 iterations: 1.487863\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/mnt/d/Xiangyu/AWARE/utils/clustering.py:167: UserWarning: The palette list has more values (8) than needed (2), which may not be intended.\n",
      "  sns.scatterplot(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 91 nearest neighbors...\n",
      "[t-SNE] Indexed 837 samples in 0.000s...\n",
      "[t-SNE] Computed neighbors for 837 samples in 0.035s...\n",
      "[t-SNE] Computed conditional probabilities for sample 837 / 837\n",
      "[t-SNE] Mean sigma: 1.924096\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 66.874878\n",
      "[t-SNE] KL divergence after 300 iterations: 1.485631\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/mnt/d/Xiangyu/AWARE/utils/clustering.py:167: UserWarning: The palette list has more values (8) than needed (2), which may not be intended.\n",
      "  sns.scatterplot(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils.clustering import visualize\n",
    "raw = np.zeros((0,84))\n",
    "for split_idx, (train_loader, test_loader) in enumerate(splitter):\n",
    "    inputs, masks, targets, info = next(iter(test_loader))\n",
    "    raw = np.concatenate((raw, inputs), axis=0)\n",
    "\n",
    "visualize(raw, y_pred==y_true, legend=['False predictions', 'True predictions'])\n",
    "\n",
    "visualize(raw, y_true, legend=['True healthy', 'True asthma'])\n",
    "\n",
    "visualize(raw, y_pred, legend=['Predicted healthy', 'Predicted asthma'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "36dd1ef6-9af4-437f-92ec-363bb758d365",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'str'>\n"
     ]
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "750d212a-0877-4a0d-986c-57eb61d3875c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}