Run_Sidarthe_ODE.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "2bb3e3b8-442c-4b75-8c1f-863f02bde29a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.734017Z",
     "iopub.status.busy": "2022-10-02T21:32:06.733681Z",
     "iopub.status.idle": "2022-10-02T21:32:06.751620Z",
     "shell.execute_reply": "2022-10-02T21:32:06.750526Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.733979Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "from covid_ode import sidartha_ode, covid_dataset\n",
    "import pandas as pd\n",
    "from torch.utils.data import DataLoader\n",
    "from tqdm.notebook import tqdm\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "92f384c5-3b0b-41ce-8688-bf9ad652fbc1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.753867Z",
     "iopub.status.busy": "2022-10-02T21:32:06.753393Z",
     "iopub.status.idle": "2022-10-02T21:32:06.819507Z",
     "shell.execute_reply": "2022-10-02T21:32:06.818695Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.753809Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "canada = pd.read_csv('canada.csv')\n",
    "canada = canada.values[:, 0:3].astype(float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "ed7cdb24-fa5d-4b71-a723-841fd3f5e7c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.822722Z",
     "iopub.status.busy": "2022-10-02T21:32:06.822409Z",
     "iopub.status.idle": "2022-10-02T21:32:06.827354Z",
     "shell.execute_reply": "2022-10-02T21:32:06.826276Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.822670Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "920e90ef-5f1a-4091-bbfd-7e546d8793f8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.830381Z",
     "iopub.status.busy": "2022-10-02T21:32:06.830146Z",
     "iopub.status.idle": "2022-10-02T21:32:06.835501Z",
     "shell.execute_reply": "2022-10-02T21:32:06.834132Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.830356Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(555, 3)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "canada.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "0489017c-94f1-41a5-b54b-5c4629f0110d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.837889Z",
     "iopub.status.busy": "2022-10-02T21:32:06.837649Z",
     "iopub.status.idle": "2022-10-02T21:32:06.844544Z",
     "shell.execute_reply": "2022-10-02T21:32:06.843580Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.837862Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[4.000000e+00, 0.000000e+00, 0.000000e+00],\n",
       "       [4.000000e+00, 0.000000e+00, 0.000000e+00],\n",
       "       [4.000000e+00, 0.000000e+00, 0.000000e+00],\n",
       "       ...,\n",
       "       [1.435683e+06, 1.404327e+06, 2.660200e+04],\n",
       "       [1.437713e+06, 1.405529e+06, 2.660800e+04],\n",
       "       [1.438806e+06, 1.405971e+06, 2.661900e+04]])"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "canada"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "5f80af06-d168-45c8-b41b-eb584131862b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.846980Z",
     "iopub.status.busy": "2022-10-02T21:32:06.846669Z",
     "iopub.status.idle": "2022-10-02T21:32:06.863154Z",
     "shell.execute_reply": "2022-10-02T21:32:06.862431Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.846951Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MinMaxScaler()"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TRAIN_IDX = int(canada.shape[0]*.5)\n",
    "VAL_IDX = TRAIN_IDX + int(canada.shape[0]*.1)\n",
    "\n",
    "min_max_scaler = MinMaxScaler()\n",
    "min_max_scaler.fit(canada[:TRAIN_IDX])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "2258e7da-b747-4cfc-a5b2-056e2603fd11",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.864671Z",
     "iopub.status.busy": "2022-10-02T21:32:06.864232Z",
     "iopub.status.idle": "2022-10-02T21:32:06.870172Z",
     "shell.execute_reply": "2022-10-02T21:32:06.868968Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.864626Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "canada_train = min_max_scaler.transform(canada[:TRAIN_IDX])\n",
    "canada_val = min_max_scaler.transform(canada[TRAIN_IDX:VAL_IDX])\n",
    "canada_test = min_max_scaler.transform(canada[VAL_IDX:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "7190f0b0-abf6-4a22-aa60-3a92d6e3ec64",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.873792Z",
     "iopub.status.busy": "2022-10-02T21:32:06.873567Z",
     "iopub.status.idle": "2022-10-02T21:32:06.880808Z",
     "shell.execute_reply": "2022-10-02T21:32:06.879169Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.873769Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Learning Rate\n",
    "lr = 0.0001\n",
    "\n",
    "# number of historic data points for fitting\n",
    "input_steps = 1\n",
    "\n",
    "# forecasting horizon\n",
    "output_steps = 2\n",
    "\n",
    "# number of epochs for training\n",
    "num_epochs = 50\n",
    "num_epochs_for_test = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "962d427e-5aee-4bb5-b63d-5710493a938f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.884110Z",
     "iopub.status.busy": "2022-10-02T21:32:06.883821Z",
     "iopub.status.idle": "2022-10-02T21:32:06.889262Z",
     "shell.execute_reply": "2022-10-02T21:32:06.887780Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.884079Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "VAL_IDX-TRAIN_IDX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "e474c655-0acf-4fbe-b182-8c67fef6a5b9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.891407Z",
     "iopub.status.busy": "2022-10-02T21:32:06.891133Z",
     "iopub.status.idle": "2022-10-02T21:32:06.900584Z",
     "shell.execute_reply": "2022-10-02T21:32:06.899466Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.891377Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "training_data = covid_dataset.CovidDataset(canada_train, input_steps, output_steps)\n",
    "train_dataloader = DataLoader(training_data, batch_size=1, shuffle=False)\n",
    "\n",
    "val_data = covid_dataset.CovidDataset(canada_val, input_steps, output_steps)\n",
    "val_dataloader = DataLoader(val_data, batch_size=1, shuffle=False)\n",
    "\n",
    "test_data = covid_dataset.CovidDataset(canada_test, input_steps, output_steps)\n",
    "test_dataloader = DataLoader(test_data, batch_size=1, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "1e335d20-9338-4a4e-8c41-9bd69d7748dd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.902550Z",
     "iopub.status.busy": "2022-10-02T21:32:06.902245Z",
     "iopub.status.idle": "2022-10-02T21:32:06.913263Z",
     "shell.execute_reply": "2022-10-02T21:32:06.912291Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.902514Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def train(dataloader, limit=None):\n",
    "    for i, data_with_exact in tqdm(enumerate(dataloader), total=limit if limit else len(train_dataloader)):\n",
    "        losses = 0\n",
    "        data, y_exact = data_with_exact\n",
    "        optimizer.zero_grad()    \n",
    "        y_approx = model(D0=data[0,0,0], H0=data[0,0,1], E0=data[0,0,2])\n",
    "        loss = loss_fun(y_approx[1:, \n",
    "                                 [model.pos_dict['D'], model.pos_dict['H'], model.pos_dict['E']]], # D, H, E\n",
    "                        y_exact.squeeze())\n",
    "        loss.backward()\n",
    "        losses += loss\n",
    "        optimizer.step()\n",
    "        scheduler.step()\n",
    "        with torch.no_grad():\n",
    "            state_dict = model.state_dict()\n",
    "            for k in state_dict.keys():\n",
    "                state_dict[k] = torch.clamp(state_dict[k], min=1e-2)\n",
    "            model.load_state_dict(state_dict)\n",
    "        if limit and i == limit:\n",
    "            break\n",
    "    print(\"Train losses:\", losses/i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "5980097f-9a1d-412d-8365-6d7fd2429dac",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:32:06.915869Z",
     "iopub.status.busy": "2022-10-02T21:32:06.915579Z",
     "iopub.status.idle": "2022-10-02T21:33:51.920619Z",
     "shell.execute_reply": "2022-10-02T21:33:51.919690Z",
     "shell.execute_reply.started": "2022-10-02T21:32:06.915831Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1016751170f0497e9ef03ba008445876",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0114e-05, grad_fn=<DivBackward0>)\n",
      "Val losses:  tensor(0.0179)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "abc2533efb264cd8bc586e35ffed88fc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2045e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8d433616ae4e4945bfa5d8b1eed67e31",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2431e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9e8d76c036b24055ac6805dfcc896bc6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2294e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a63d05a6ff3429db99174c54826a1d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1483e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "21279bb81cee40fb8d0f8a35248b42c3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0538e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "12a034eaec0145fb85274127e59038f1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.5467e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "85b72aa85b704ce19d8957ab664d7e7c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(8.7583e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1070450b3932477a9437c7b99f64d00a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.8915e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a580870adb7d4fafb8d04784143ca145",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.0654e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "018c9b4f6c70474eb2a913f4d7dd3628",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(6.3586e-06, grad_fn=<DivBackward0>)\n",
      "Val losses:  tensor(0.0090)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9fab788d3ef348248830a7da5e498cc6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(5.7203e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0dc956a8f37e4bfea07b97ba3d040002",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(5.1115e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "97d498d9387c4db3bdb42c25c0789fba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.5398e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "45d8baaddcaf42fd9774f18090e00ff3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.1204e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1029f46b38654b5ab8f32ccd2e5e724e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.6593e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "74fa4abdf7964078a09311f463cd0584",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.2375e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "06b693bd31af4dfca34f72f950390d7d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.8517e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0cb41732136840358de786b116f14c05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.5537e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1857b52b80e940e4a0f711527b7c7320",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.2656e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "475404c732644c8c8c225ac280505804",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9986e-06, grad_fn=<DivBackward0>)\n",
      "Val losses:  tensor(0.0029)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3465ed6a38654c818cb6e69f37d4d360",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7945e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "27d21272d4964a61951774e8098870ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5885e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b79b2220899441ebbb3b7a52d31de5a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3999e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "47829a53d5ad4fe3aa66af5d1e444dae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2282e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "734e7d234bde4d42a9ecfc83d87ce94f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1018e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a885b4a77f024c8b831f771cb1dcfedb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.7107e-07, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3cac7cabceec4b8f917c0c9faf621f2b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.4974e-07, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6d4fb29ef0cb4f0a8d717427637cea41",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0256e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3092da41d7a34f35b292443671ccb2b3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0804e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e202be4ac00049dc80544100029a70fb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1122e-06, grad_fn=<DivBackward0>)\n",
      "Val losses:  tensor(0.0016)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b16dd4644ae44c31b0c281487c650d49",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1315e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d1463a8278c44ae9a97bf5a989ee9c84",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1500e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "77595b6ceb7e493c9ae12de054bb731d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1535e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65372561aa39491ba05ae4c42b329746",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1534e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6f180d8d9da64dc69f55756aba114c03",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1511e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58955172069f4e3a820fe6697232be82",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1563e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7958bd4eb9f743ef9264c524edbe56a4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1514e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ef192d1902904725bc8945935e88bf67",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1463e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "08edde36d32f4767885b95991040537e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1417e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "163e46545a3a4d9fb6f538e43012a549",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1419e-06, grad_fn=<DivBackward0>)\n",
      "Val losses:  tensor(0.0016)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "121b0b545f2b40a6b0e02b21b9fee72d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1387e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3cc96b13566b4edeb3728dc388291e29",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1353e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c686ea9e176411c8ce649245aa79992",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1394e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "72e61309969e44f48f976212f91afadb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1356e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f3018da95d1445ce938aa1a60f04cfd6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1319e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0fe145b591a44c5b85724e83a12b5e39",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1282e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "748b9dbad3c2487a88d4baf697c3b26f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1312e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8fb9046b9868485fbb63ff5884da5b29",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1276e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "06b19797c5c04a7bb272f3ea959f5a7e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1240e-06, grad_fn=<DivBackward0>)\n"
     ]
    }
   ],
   "source": [
    "model = sidartha_ode.SIDARTHEOde(len_data = output_steps).to(device)\n",
    "\n",
    "torch.manual_seed(3407)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr)\n",
    "scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size= 1000, gamma=0.9)\n",
    "loss_fun = torch.nn.MSELoss()\n",
    "\n",
    "for _ in range(num_epochs):\n",
    "    train(train_dataloader)\n",
    "    if (_ % 10) == 0:\n",
    "        validation_prediction = []\n",
    "        with torch.no_grad():\n",
    "            val_losses = 0\n",
    "            for data, y_exact in val_dataloader:\n",
    "                y_approx = model(D0=data[0,0,0], H0=data[0,0,1], E0=data[0,0,2])\n",
    "                validation_prediction.append(y_approx.numpy())\n",
    "                loss = loss_fun(y_approx[1:, \n",
    "                                     [model.pos_dict['D'], model.pos_dict['H'], model.pos_dict['E']]], # I, H, E\n",
    "                            y_exact.squeeze())\n",
    "                val_losses += loss\n",
    "            print(\"Val losses: \", val_losses/len(val_data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "339f5a67-fed4-4d49-a3b5-7fe3371c69f4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.922375Z",
     "iopub.status.busy": "2022-10-02T21:33:51.921999Z",
     "iopub.status.idle": "2022-10-02T21:33:51.926591Z",
     "shell.execute_reply": "2022-10-02T21:33:51.925702Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.922329Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# predict, confirmed, recovered, death\n",
    "validation_prediction = np.array(validation_prediction)\n",
    "observerable=[validation_prediction[:, :, [model.pos_dict['D'], model.pos_dict['H'], model.pos_dict['E']]]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "f6dd5928-b7a3-489c-b8e6-6c6ad9a9c631",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.929170Z",
     "iopub.status.busy": "2022-10-02T21:33:51.928823Z",
     "iopub.status.idle": "2022-10-02T21:33:51.937165Z",
     "shell.execute_reply": "2022-10-02T21:33:51.936186Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.929113Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "inversed_predict = [min_max_scaler.inverse_transform(i) for i in observerable[0]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "0e1b993d-8696-469d-b7b7-adab503fff07",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.938738Z",
     "iopub.status.busy": "2022-10-02T21:33:51.938354Z",
     "iopub.status.idle": "2022-10-02T21:33:51.943526Z",
     "shell.execute_reply": "2022-10-02T21:33:51.942246Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.938691Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "inversed_predict = np.array(inversed_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "956a1b7e-4d30-45b0-870c-0f0210b18145",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.945514Z",
     "iopub.status.busy": "2022-10-02T21:33:51.945240Z",
     "iopub.status.idle": "2022-10-02T21:33:51.949754Z",
     "shell.execute_reply": "2022-10-02T21:33:51.948561Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.945488Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "predict_1_ahead = inversed_predict[:,1]\n",
    "predict_2_ahead = inversed_predict[:,2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "abfe8c27-62ee-42c5-9f8b-56bb05394e0f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.952003Z",
     "iopub.status.busy": "2022-10-02T21:33:51.951712Z",
     "iopub.status.idle": "2022-10-02T21:33:51.956181Z",
     "shell.execute_reply": "2022-10-02T21:33:51.954986Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.951972Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "true_value=min_max_scaler.inverse_transform(canada_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "2839523b-190a-43c7-bfc2-ee93b68c0a3d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:51.958688Z",
     "iopub.status.busy": "2022-10-02T21:33:51.958090Z",
     "iopub.status.idle": "2022-10-02T21:33:52.117519Z",
     "shell.execute_reply": "2022-10-02T21:33:52.116426Z",
     "shell.execute_reply.started": "2022-10-02T21:33:51.958638Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "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 matplotlib.pyplot as plt\n",
    "\n",
    "x = range(len(true_value))\n",
    "\n",
    "plt.plot(x[3:], predict_1_ahead[:,0], label ='predict infected 1')\n",
    "plt.plot(x, true_value[:,0], '-.', label ='actual infected')\n",
    "plt.plot(x[3:], predict_1_ahead[:,1], label ='predict recovered')\n",
    "plt.plot(x, true_value[:,1], '-.', label ='actual recovered')\n",
    "\n",
    "plt.xlabel(\"date\")\n",
    "plt.ylabel(\"number of people\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "d732d879-c924-4a65-802c-f76c8a12ad2d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:52.119077Z",
     "iopub.status.busy": "2022-10-02T21:33:52.118843Z",
     "iopub.status.idle": "2022-10-02T21:33:52.244794Z",
     "shell.execute_reply": "2022-10-02T21:33:52.244134Z",
     "shell.execute_reply.started": "2022-10-02T21:33:52.119050Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "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": [
    "plt.plot(x[3:], predict_1_ahead[:,2], label ='predict death')\n",
    "plt.plot(x, true_value[:,2], '-.', label ='actual death')\n",
    "plt.xlabel(\"date\")\n",
    "plt.ylabel(\"number of people\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "efbccded-7151-4596-9458-4b16e9e32a83",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:52.246398Z",
     "iopub.status.busy": "2022-10-02T21:33:52.246167Z",
     "iopub.status.idle": "2022-10-02T21:33:52.252138Z",
     "shell.execute_reply": "2022-10-02T21:33:52.251261Z",
     "shell.execute_reply.started": "2022-10-02T21:33:52.246371Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21788.12751024525"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_ahead[:,0], true_value[3:,0], squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "6d3887a8-b04e-489d-9db5-2e0bf3d9d568",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:52.254586Z",
     "iopub.status.busy": "2022-10-02T21:33:52.254054Z",
     "iopub.status.idle": "2022-10-02T21:33:52.261297Z",
     "shell.execute_reply": "2022-10-02T21:33:52.260424Z",
     "shell.execute_reply.started": "2022-10-02T21:33:52.254520Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10498.449974516981"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_ahead[:,1], true_value[3:,1], squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "25ca5b5f-0f27-46da-9361-5f2b492de7a8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T21:33:52.263298Z",
     "iopub.status.busy": "2022-10-02T21:33:52.262660Z",
     "iopub.status.idle": "2022-10-02T21:33:52.269439Z",
     "shell.execute_reply": "2022-10-02T21:33:52.268472Z",
     "shell.execute_reply.started": "2022-10-02T21:33:52.263257Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "241.75379541962772"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_ahead[:,2], true_value[3:,2], squared=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "efeeba88-9b7b-41b6-a556-1edd9403ccf4",
   "metadata": {},
   "source": [
    "### Predict whole test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "8a348aef-bee7-478f-b7b8-417737651488",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:27:14.391016Z",
     "iopub.status.busy": "2022-10-02T23:27:14.390628Z",
     "iopub.status.idle": "2022-10-02T23:27:14.882909Z",
     "shell.execute_reply": "2022-10-02T23:27:14.881859Z",
     "shell.execute_reply.started": "2022-10-02T23:27:14.390960Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec15c7727b504a5db5c16cbb1e1ab2dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/52 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(5.3822e-05, grad_fn=<DivBackward0>)\n"
     ]
    }
   ],
   "source": [
    "train(val_dataloader, limit=len(val_dataloader))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "3dfaadfe-6d89-4d00-9475-e0bc48645d23",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:27:15.513393Z",
     "iopub.status.busy": "2022-10-02T23:27:15.513031Z",
     "iopub.status.idle": "2022-10-02T23:30:42.615816Z",
     "shell.execute_reply": "2022-10-02T23:30:42.615252Z",
     "shell.execute_reply.started": "2022-10-02T23:27:15.513350Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9be9f992d1724a19a3984fea622631ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/220 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "701c6ebc99654241ad7d2cf7f0a55580",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.9180e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6dd21216c56a4551b71832ad95c1cc3b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.5368e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "66025b3670bc427da88e870c4085de8c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/274 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.2393e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a4cce46092145b2968321f4eeff2d72",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/3 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0008, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d15bf5a5c2d94c9d9fd63a14c95b24e6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/3 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0008, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c4192f014d84d4fba0297f37395b77b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/3 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0008, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "606587bc6f54482285b6a19e9c5d11c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5e02601935c64288befcf3d78440cd80",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bd6dc2126a9e49089370b82691b7f071",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7ef7fa0c966045e6922c99bb8638eea0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "eae6e0484fc14ed8874a636deb7212fa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "81759ec8a38c40e981857cb22af36f93",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0003, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4662214f142d4fc9baf73c9c7400f28a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/12 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0002, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e1f62d0e589349769a6220a89210d944",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/12 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0002, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9e62a8de9ac845dabe61e338fb294c57",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/12 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0002, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "156577850c984248b67489101c26035e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/15 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4e62f64316b34ed3a5cf4b24d15b5f4e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/15 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8bdb719607934a018b66349436b288ab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/15 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1e4d2ba53d246069e2378b6c45da753",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/18 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "652a8539a2f14e6b91f4fd5f23853444",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/18 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ba1afa06f8644790b7f2bb66774f2939",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/18 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(0.0001, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d120a5e6415e483e8ae94888525ee638",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/21 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.7996e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f5ed5b677b1646a2a5e1844940b1935c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/21 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.7059e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "55fccea63c8e4b5791d5a36ce8e4248a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/21 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.6127e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ff498692bc8744109d9efa3105279304",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/24 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.2465e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e8ae1ec717b44f50aa4380c3be2f44e8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/24 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.1467e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d908aa17909547cb94dd7e7e37bcc677",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/24 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(7.0474e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8adfa150bb0746758ceccde920d13019",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/27 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(6.2610e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9ff22ecf69ff44b0956c351a1d38ee6d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/27 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(6.1676e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "62f44242028c4a67861ae8ecb9968fb1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/27 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(6.0749e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "404ae15063c644b291b7bf2b43f0e965",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/30 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(5.1653e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "99c2776f3494414c9f7a12956d8642fd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/30 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(5.0730e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "202b735844f942dc85f256a79d2d6fd9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/30 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.9814e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "224e5c00f9d74fdfab13b69049cfc1a0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/33 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.9408e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "04f4c662635b4c3897c5b6ff71b55e88",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/33 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.8498e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f84e0abd60e546ba9f807d92acfa2e3d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/33 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(4.7675e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "38454c5cd83c44b7a6f72d9b71871688",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/36 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.3376e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "894f7aa0f959483eaf97187d34fe046c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/36 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.2675e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9176f59483f641b1bce8cae7e15a884a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/36 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.1980e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c5c84d1500114d7daf244e254446e4fb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/39 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.8688e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "abade4affd264358b6e340da1dde2fff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/39 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.7911e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "696f0e31b79a4eb99aa22f897b4b08ee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/39 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.7140e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1595bb4ae4d148bca7ef774c3551b6f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/42 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.2038e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6b5fb9278ab24214b2db514a820c3e78",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/42 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.1430e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8859c5d8911a4a6e8eb3c1058110fcc7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/42 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.0829e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "01a28bf49b304cf8b3473fd2c4b08536",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/45 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.7854e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2430d1ced2f42059b8f02bdbc2d7f69",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/45 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.7233e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d5de5ac45da248d5bcd42866a9648570",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/45 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.6622e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c8135ecab3f7440bb4cd74682bd1f4f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/48 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.0832e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9d9c538c35c4279a769113ed357a915",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/48 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(3.0228e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9210e80ba2f840049da00b95ba77f97f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/48 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.9635e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63a5164d3f8f4a06aa196d71d8c5b801",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/51 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9879e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "26a5b038ee924227be0ad4e2c6ecd363",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/51 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9365e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ac0fc89dc5834d7c92a7070bdc2ae505",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/51 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8871e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e42e79d5aeea47a8ac9af693b61d32d6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/54 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.3664e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aae96e6f964d4922a62a636da500132f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/54 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.3132e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1a3c9629f08c4db8ae8048f656a458dc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/54 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.2615e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bbace26da3764a9987742be23bc9e3f1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/57 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.2307e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "023e6dcbd24b4288a09d2abcb55e8bc3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/57 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.1886e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5744f27e9cce4e00b207e6fe9ff7b8a2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/57 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.1475e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "730329e29bd04b1493bad76da8bad985",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/60 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.1294e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f2867890bd964cb2b05de8e1e563cc32",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/60 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0872e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f8c94303fac6446d8f5088a850eb4039",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/60 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0461e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cab13d5f8e594f47bed3d501ca740103",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/63 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0523e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7580c537428b48b5b87a47321c943126",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/63 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0175e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ea7e087e7b024c389126758ff699caa8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/63 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9838e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b49e16f3752641ea855f32dffb662455",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/66 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8504e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "68cffba9c82a4455b0dea1097084cd01",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/66 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8130e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d5c00e1b08a644c8bf1fdf3efb242111",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/66 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7767e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e3e6ec43173a420f9e3541014b22600c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/69 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8173e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7f1ecd7624e8438eaba5193c667ffbd8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/69 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7819e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3fee6fdca8154c1f87bfbca8ced7fcd8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/69 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7478e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "093bac701fff487da0795440447d44a7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/72 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5027e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b752bc3df39b42019c83ddee8baf84f6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/72 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4702e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "85e473ad935a4105847426a2f6b7633e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/72 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4389e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2867f423a26347ee9d89c7f3ad8aedfd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/75 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6585e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2899b707ac354578b054170e45f19fdf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/75 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6278e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4a60f42d61914750b521c6b3c116338a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/75 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5983e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "79095f6515c04b72aec9ef9f9cfb08f9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/78 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4557e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f1116912e5a04572b9e7085b38225dd4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/78 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4306e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4e722fcfc4943eca0d0d93ed2e88f2d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/78 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4068e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8926aec35abf4dea917f2fa8af529f68",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/81 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4809e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d8637cb14fea487b84b294c1d4c2d2d9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/81 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4538e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "485a29090224456a83d3310828fe50ff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/81 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4279e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "87b9a11fbbd6463683a2386f50635d2c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/84 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5159e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "47cd0cb42475470dbaafb23c17d29f27",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/84 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4953e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "38b7793023d547c2834f26a1eb6114f9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/84 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4765e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65697da57ee041db94e70d29abb1d01b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/87 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4891e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "613088a38691459b9135127e43b4325b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/87 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4648e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "515dd6561a5f46e18ec29cd8db20706e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/87 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4418e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07106738725344d8a7f27e81c848cfe0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/90 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4685e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8fe7d1fe1d544881abdc73a04cbb43f1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/90 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4430e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d00ff549a2104c20a80c1ed726a4f668",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/90 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4189e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9a03f68e29c46ed8554b83d7ac2ff8f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/93 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4190e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e0ad132a681944afaba6f966bce436f6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/93 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3907e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d8598fc090f8472a9edc6a0b83d826e3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/93 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3641e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "873da57c24d547afbe26657d037ac8e0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/96 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5942e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9096f4b1a2dd429282ccad7c4a4f2039",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/96 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5718e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b2249d4947bc4bbd983a06ea55ee2727",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/96 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5509e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e566b1a2fdf74c6fa8d9acd3d1f830a8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/99 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4208e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f06b6b1950a48e0a4f04a6a72746155",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/99 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4001e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb63efa8ca054259a1cbd55e5af33f0f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/99 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3809e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "68f59eb888ad43f7aeaeff809e8e6353",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/102 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5485e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f60dba5d7972481f95f43753dc5d0512",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/102 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5178e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d69d6c40878f4ba996a0a2fa248ff9ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/102 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4899e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d002d56c0a8946a5b6ab6e54c90ec4c3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/105 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5244e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "39ecd460853a4eb79087b11b43f178b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/105 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4991e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ce8d0350cb6e446892b9ab117887394d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/105 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4775e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a16c8260c4b0484daac34990b46c36fb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/108 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4780e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6cb11d75ac32429c86782998c9fee471",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/108 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4521e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "38827f2cd10c497e80e0bf3f036660f7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/108 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4289e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec57020242bf41d387d98c5ff3a1cbba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/111 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2594e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f374cefebe9d4257b5dc3bb5bdfb8c4a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/111 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2363e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "905c4d236bc6421280bb5368f91aab8d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/111 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2156e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e4b661d12f594393b438e87e59c5a28c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/114 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2492e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "de2f4f971d8045ec866cb865eea5edd9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/114 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2269e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c0e07193aab7470ab477673602029250",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/114 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2080e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6964e51553624ba090730357c2239ac9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/117 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1720e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1bcb9f89334a497b8293912b5b8540da",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/117 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1523e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3f5b89e39b4a4829bae75b3033ed5623",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/117 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1344e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c5a2309fc0c46e19c8c5abec4a67978",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/120 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0778e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ee4f1e056fee4ba4898b961448ad8195",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/120 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0617e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db1ea5e069454f09bea4e80982510028",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/120 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0471e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f740d17fc17f415194fdb4dfa71db13e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/123 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2202e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7169d3ea7af947a4aabddf98455ebcad",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/123 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2045e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6ec50e4a73ea4fb5b1cdf589d76cc642",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/123 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1901e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "da92f15163d048ab879ca549dd4e3657",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/126 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4352e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "195261fb0d834e16949cc19201a34637",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/126 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4215e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "09c20b811bb94cadb3c611aa01be7a9d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/126 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4089e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ab590ff9719c410b8169f108af1fd798",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/129 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1608e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f47698be2acf4d909f938adad36754de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/129 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1472e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f5a7ec73c5b147e7838dc43f77dac0b7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/129 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1364e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e831bc51d404a9880e378160d1d708d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/132 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3640e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a5ecef01c52647f79850f87919d7f7d7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/132 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3515e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e89c1496566b4920a9c71da803ce2f82",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/132 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.3400e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1d8b22c10c894a31a4f240aabfe4f951",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/135 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.4914e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cd0f621cc8ba4e8987c1a7369f0fa19a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/135 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.3779e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4ba4ffe04de54a5ea14a1e517b9929e2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/135 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.2722e-06, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8fabe40bc90a46e48abc5358c0e092e5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/138 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1346e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "768df887f79c44c4af0ea97ecf1c99bc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/138 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1250e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "66629c48f8064d2a9621bd440eeaaca4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/138 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1161e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f2cc5adef9b416ca86e3121db89bfe1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/141 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.5966e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "93c76a78b117474dace3474c7b91c3da",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/141 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.5214e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4033020cfb8746f2b47450b92c42f40c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/141 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.4508e-06, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1c2675ac29674aa2bd21cc85f4387139",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/144 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0567e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f9138558f95d49e9afbd6d2dae1e4494",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/144 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0490e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "85cc9e5b62304462a9b2b0d9b75ecd0b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/144 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0425e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51506ad8ea9a4a8786786b44d3e2ef42",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/147 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4806e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "543104fbb0c9480fa21c71e8ea6d0000",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/147 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4755e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f971e04f9bd244ddb8096789558006cd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/147 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.4706e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8127c606d5d54a77833e130dda1773de",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/150 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.4454e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c08c63a2b7a34c30b56cb3558139f6ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/150 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.3962e-06, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c24577068dfd452f98d71379bff72a29",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/150 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(9.3581e-06, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "811136ebd4744e07a916d7e0aaed63f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/153 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2616e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6c727c5fc55f4833be1018a14a750664",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/153 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2577e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ca73de8c45134aafb07209f5c60a7bcf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/153 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.2539e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ca45a39fd95945d2a77d94b26cc2c90a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/156 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1039e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6fee062022e1489aacd582f3846c6151",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/156 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.1006e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2c3e1a765215487c896a79883434a89d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/156 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.0974e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "41a97d6074ca4a638b4ba2f1431dd739",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/159 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5389e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4ce4ab55229b441783448a2b3499f45d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/159 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5361e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4576c547a6b0465785cf79d053a6cec4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/159 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.5335e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "339b446817b34788bb83c32d658d1132",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/162 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6094e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "da9c3f174e6147f5893d40a05f63c5a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/162 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6072e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7cd61d604fde42d7a53ef438321e1a67",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/162 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6050e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2375352ab16f4b5c9fa1408d80ffa6cd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/165 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6661e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5b07cbf3374240fabd4602abeabaa5be",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/165 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6641e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e7624744a6164838bf32e6dd838b0c64",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/165 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.6622e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "913d40126b864194b6783d24f1fe2cb1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/168 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8521e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e0b36fe84c64d92b1c696001367beb1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/168 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8504e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65e06eb31d214171985a1cf8d19eb0a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/168 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8487e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a16a2c611d74933b579518229569124",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/171 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7813e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "de1087a40f294899b2cd1be6aaf7f3a2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/171 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7799e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d7f49b9fbec3478a9c82c3030bcbdd67",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/171 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7785e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f6940fc359ae4604a08ba1a145307209",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/174 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8832e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "656a797bcfbb4683b97ef0756c1c6d15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/174 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8818e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "02a1b45d5dc848e7b959b530f477a85d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/174 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8805e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "eae6f0c287744f698a73879ea9fb4bc7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/177 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7471e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e53995ea5734770bd9a8e7d29601e42",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/177 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7461e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db0e06db7253430f8f75dd1a4292a246",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/177 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.7450e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c9aabc63c22248eba3bce209b49c3c59",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/180 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9512e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "755c9ab61c15466cb3d2ac2d4bcc1bff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/180 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9501e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "61cfc3d00f34433197386d206db878a0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/180 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9493e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3e61d06892d64952b9e0240bf23e2782",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/183 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8552e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e560d6cc000a44cbb6724af46e9ef28c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/183 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8543e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "898d34f7d10444e9922b0c790b505c1b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/183 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8534e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2cd916d0b712485890895f13462d4f46",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/186 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9746e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a8df1fc4d8a4818a86e1c7ed5103ae4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/186 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9737e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4e401879f25d427eb1ef4ee93b7aad6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/186 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9730e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5474ae49150646f895db8d91f1396b97",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/189 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0220e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "06b4f9b543e744369966a60932756adb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/189 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0212e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "42b2a7fbc6ff4ffebafd414f4b80fd93",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/189 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(2.0205e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "43f8ea4d22d046a197397c19ae1a2b4a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/192 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9225e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "32495c88daad4b5795a41e2212862058",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/192 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9219e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e067b7c3afe24bc395bb4f5f5b6d905c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/192 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9213e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "be8159f14c7a48bab1d5dab944706c63",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/195 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8930e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3ca0c3a99e7f4ee89e9007cdc66d51e6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/195 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8923e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6dd727d6b9bf47b2b89671bbfdef56b8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/195 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8917e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c811e7ac62a1432ebbb741dd31a7d1e1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/198 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8353e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c42cb2e743f947b2a8152eaaba1da354",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/198 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8347e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "90e14e3ad0e8417ba5bd28afd06fcec5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/198 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8342e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c7dc5369c851471cb4e16d247eaacc05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/201 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9150e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "66d6e1dad904463dad6253f2641898fa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/201 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9144e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "af5bf7a636a84ff2afcff5697eb9f832",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/201 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.9140e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8d76f40b0bac4745a534147f5cd0ecdf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/204 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8706e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "38264cc28c774c52974c721704885dd4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/204 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8701e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "146f651eb4cf42b0b0ccf42f1bac06df",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/204 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8696e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "71bbb1e0588a4368a67b3ffffb5979a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/207 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8348e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5a9e51b8436248ffbc4d3e5c8d04dc2a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/207 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8344e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ea3d7682321247e5bd3e8ac45856a249",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/207 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8339e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3c3b7253d2b2498797b940b60a59bb4a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/210 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8868e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f6e0e4e1f6c488ca3379ed9ea1c1fc2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/210 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8864e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c06d378ee6634672a6b78e046db00562",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/210 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8859e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "559cc32895fc4a85be337e0a44409dfe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/213 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8177e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3c9d27124dff4556b5c2686ac2ead6b8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/213 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8172e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7efa1bf10be04a3a8553d9fccf65b732",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/213 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8168e-05, grad_fn=<DivBackward0>)\n",
      "Retrain on past test\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1c898a3d1c0740f6a3ed16498a424e29",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/216 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8251e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2757230281594e61b75f2777157f658c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/216 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8247e-05, grad_fn=<DivBackward0>)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b5bf8c3fe6ae4d43a9dc883267a808d2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/216 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train losses: tensor(1.8242e-05, grad_fn=<DivBackward0>)\n"
     ]
    }
   ],
   "source": [
    "test_prediction = []\n",
    "model_params = []\n",
    "for i, data_and_output in tqdm(enumerate(test_dataloader), total=len(test_dataloader)):\n",
    "    data, _ = data_and_output\n",
    "    # use test set prior to this step to retrain model\n",
    "    if i % (output_steps+input_steps) == 0 and i > 0:\n",
    "        print(\"Retrain on past test\")\n",
    "        for epoch in range(num_epochs_for_test):\n",
    "            train(test_dataloader, limit=i-output_steps-1)  # only train on 1-output_steps-1 sample\n",
    "    # predict\n",
    "    with torch.no_grad():\n",
    "        y_approx = model(D0=data[0,0,0], H0=data[0,0,1], E0=data[0,0,2])\n",
    "        test_prediction.append(y_approx[1:,\n",
    "                                        [model.pos_dict['D'], model.pos_dict['H'], model.pos_dict['E']]].detach().numpy())\n",
    "        \n",
    "    model_params.append([x.data for x in list(model.parameters())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "f6afcc3a-8eec-4bf6-bf04-db92db870996",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.617415Z",
     "iopub.status.busy": "2022-10-02T23:30:42.617195Z",
     "iopub.status.idle": "2022-10-02T23:30:42.620337Z",
     "shell.execute_reply": "2022-10-02T23:30:42.619542Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.617391Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# test_prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "efb8754f-c84d-412e-b74e-88fe55d08022",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.623763Z",
     "iopub.status.busy": "2022-10-02T23:30:42.623175Z",
     "iopub.status.idle": "2022-10-02T23:30:42.628823Z",
     "shell.execute_reply": "2022-10-02T23:30:42.627647Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.623730Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "test_prediction = np.array(test_prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "f74b6e54-7c4c-4bbb-ae02-e54edfbdff09",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.631353Z",
     "iopub.status.busy": "2022-10-02T23:30:42.630899Z",
     "iopub.status.idle": "2022-10-02T23:30:42.635299Z",
     "shell.execute_reply": "2022-10-02T23:30:42.634372Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.631305Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "predict_1_day_ahead = test_prediction[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "8603e019-04da-4162-bb06-f7b4d595187d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.637409Z",
     "iopub.status.busy": "2022-10-02T23:30:42.637154Z",
     "iopub.status.idle": "2022-10-02T23:30:42.642204Z",
     "shell.execute_reply": "2022-10-02T23:30:42.640856Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.637375Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "predict_1_day_ahead_observable=min_max_scaler.inverse_transform(predict_1_day_ahead)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "0919f670-b413-4084-b1df-be4c8e990180",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.644491Z",
     "iopub.status.busy": "2022-10-02T23:30:42.644163Z",
     "iopub.status.idle": "2022-10-02T23:30:42.648934Z",
     "shell.execute_reply": "2022-10-02T23:30:42.647827Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.644438Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "true_test_value = min_max_scaler.inverse_transform(canada_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "51425000-a949-43a0-ad13-16a8330052f1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.650795Z",
     "iopub.status.busy": "2022-10-02T23:30:42.650570Z",
     "iopub.status.idle": "2022-10-02T23:30:42.664780Z",
     "shell.execute_reply": "2022-10-02T23:30:42.663753Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.650771Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "\n",
    "canada_csv = pd.read_csv('canada.csv')\n",
    "date_series = pd.to_datetime(canada_csv['Last_Update'])\n",
    "last_update_pydatetime = [datetime(year=x.year, month=x.month, day=x.day, hour=x.hour, minute=x.minute, second=x.second)\n",
    "                          for x in date_series]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "4693cc05-0d87-4f80-8ea9-43701242f7ef",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:42.668469Z",
     "iopub.status.busy": "2022-10-02T23:30:42.668179Z",
     "iopub.status.idle": "2022-10-02T23:30:43.071053Z",
     "shell.execute_reply": "2022-10-02T23:30:43.070442Z",
     "shell.execute_reply.started": "2022-10-02T23:30:42.668448Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "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 tikzplotlib\n",
    "\n",
    "x = last_update_pydatetime[VAL_IDX+input_steps+output_steps:]\n",
    "\n",
    "confirmed_color=(217/256,95/256,2/256)\n",
    "plt.plot(x, predict_1_day_ahead_observable[:,0], label ='predict infected', color=confirmed_color )\n",
    "plt.plot(x, true_test_value[3:,0], '-.', label ='actual infected', color =confirmed_color)\n",
    "healed_color=(27/256,158/256,119/256)\n",
    "plt.plot(x, predict_1_day_ahead_observable[:,1], label ='predict healed', color=healed_color)\n",
    "plt.plot(x, true_test_value[3:,1], '-.', label ='actual healed', color=healed_color)\n",
    "plt.xlabel(\"date\")\n",
    "plt.ylabel(\"number of people\")\n",
    "death_color=(117/256,112/236,179/256)\n",
    "plt.plot(x, predict_1_day_ahead_observable[:,2], label ='predict death', color=death_color)\n",
    "plt.plot(x, true_test_value[3:,2], '-.', label ='actual death', color=death_color)\n",
    "plt.yscale('log')\n",
    "plt.legend()\n",
    "plt.xticks(rotation=10)\n",
    "tikzplotlib.save(\"sidarthe.tikz\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "48d3cce7-f8b3-4d09-9908-c296f68a7581",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:43.073031Z",
     "iopub.status.busy": "2022-10-02T23:30:43.072818Z",
     "iopub.status.idle": "2022-10-02T23:30:43.078976Z",
     "shell.execute_reply": "2022-10-02T23:30:43.077816Z",
     "shell.execute_reply.started": "2022-10-02T23:30:43.073008Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16410.667439629255"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_day_ahead_observable[:,0], true_test_value[3:,0], squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "36b7ec71-3aa4-4321-9311-e31407c6b231",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:43.080739Z",
     "iopub.status.busy": "2022-10-02T23:30:43.080423Z",
     "iopub.status.idle": "2022-10-02T23:30:43.087479Z",
     "shell.execute_reply": "2022-10-02T23:30:43.086106Z",
     "shell.execute_reply.started": "2022-10-02T23:30:43.080712Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10267.87672877038"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_day_ahead_observable[:,1], true_test_value[3:,1], squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "754aaf44-0ac1-40c0-91f1-d41c9d8a1db8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-02T23:30:43.089866Z",
     "iopub.status.busy": "2022-10-02T23:30:43.089408Z",
     "iopub.status.idle": "2022-10-02T23:30:43.096587Z",
     "shell.execute_reply": "2022-10-02T23:30:43.095173Z",
     "shell.execute_reply.started": "2022-10-02T23:30:43.089822Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "168.56754911020718"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(predict_1_day_ahead_observable[:,2], true_test_value[3:,2], squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "8b211522-42eb-45dc-8fb3-abb3f29aa4af",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-03T00:21:49.209653Z",
     "iopub.status.busy": "2022-10-03T00:21:49.209316Z",
     "iopub.status.idle": "2022-10-03T00:21:49.221732Z",
     "shell.execute_reply": "2022-10-03T00:21:49.220608Z",
     "shell.execute_reply.started": "2022-10-03T00:21:49.209616Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('y0',\n",
       "  Parameter containing:\n",
       "  tensor([0.5062, 0.4243, 0.0100, 0.0100, 0.0455], requires_grad=True)),\n",
       " ('α',\n",
       "  Parameter containing:\n",
       "  tensor([0.4497], requires_grad=True)),\n",
       " ('β',\n",
       "  Parameter containing:\n",
       "  tensor([0.3382], requires_grad=True)),\n",
       " ('γ',\n",
       "  Parameter containing:\n",
       "  tensor([0.4148], requires_grad=True)),\n",
       " ('δ',\n",
       "  Parameter containing:\n",
       "  tensor([0.1402], requires_grad=True)),\n",
       " ('ε',\n",
       "  Parameter containing:\n",
       "  tensor([0.1708], requires_grad=True)),\n",
       " ('θ',\n",
       "  Parameter containing:\n",
       "  tensor([0.4218], requires_grad=True)),\n",
       " ('ζ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0863], requires_grad=True)),\n",
       " ('η',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True)),\n",
       " ('μ',\n",
       "  Parameter containing:\n",
       "  tensor([0.1974], requires_grad=True)),\n",
       " ('ν',\n",
       "  Parameter containing:\n",
       "  tensor([0.1000], requires_grad=True)),\n",
       " ('τ',\n",
       "  Parameter containing:\n",
       "  tensor([0.1013], requires_grad=True)),\n",
       " ('λ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True)),\n",
       " ('κ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True)),\n",
       " ('ξ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True)),\n",
       " ('ρ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True)),\n",
       " ('σ',\n",
       "  Parameter containing:\n",
       "  tensor([0.0100], requires_grad=True))]"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(model.named_parameters())"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}