2_Training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0rFd2ZvhPWBb"
   },
   "source": [
    "# Training:\n",
    "\n",
    "This project uses a machine learning model of classification and segmentation. It is based on a U-Net architecture, which uses a set of images and binary masks to train. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build images and masks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The set of images and binary masks used for training can be generated via **satproc_extract_chips**.  \n",
    "This tool uses an image (or a set of images) and a vectorial file to create another set of images of a specific size and their masks.\n",
    "In addition, several features can be included like the rescale option to bring the output with pixel values between 0 to 255. All of them are explained below. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Obs:** If the vector file contains many categories of objects, the mask will be an image of N channels, where each one is a binary image, showing the object of interest with value one and zero otherwise. . \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![chips](img/mask_ej.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install pysatproc "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1_Pre-process.ipynb  4_Post-process.ipynb  README.md  result_model.txt\r\n",
      "2_Training.ipynb     AUTHORS.md\t\t   data       sonar-project.properties\r\n",
      "3_Prediction.ipynb   LICENSE.txt\t   img\t      tiles.vrt\r\n"
     ]
    }
   ],
   "source": [
    "!ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  0%|                                                     | 0/1 [00:00<?, ?it/s]\n",
      "  0%|                                                  | 0/1415 [00:00<?, ?it/s]\u001b[A\n",
      "  0%|                                          | 1/1415 [00:00<14:39,  1.61it/s]\u001b[A\n",
      "  0%|                                          | 2/1415 [00:01<14:11,  1.66it/s]\u001b[A\n",
      "  0%|                                          | 3/1415 [00:01<13:23,  1.76it/s]\u001b[A\n",
      "  0%|                                          | 4/1415 [00:02<13:11,  1.78it/s]\u001b[A\n",
      "  0%|▏                                         | 5/1415 [00:02<12:59,  1.81it/s]\u001b[A\n",
      "  0%|▏                                         | 6/1415 [00:03<12:52,  1.82it/s]\u001b[A\n",
      "  0%|▏                                         | 7/1415 [00:03<12:55,  1.82it/s]\u001b[A\n",
      "  1%|▏                                         | 8/1415 [00:04<13:14,  1.77it/s]\u001b[A\n",
      "  1%|▎                                         | 9/1415 [00:05<13:26,  1.74it/s]\u001b[A\n",
      "  1%|▎                                        | 10/1415 [00:05<13:09,  1.78it/s]\u001b[A\n",
      "  1%|▎                                        | 11/1415 [00:06<12:48,  1.83it/s]\u001b[A\n",
      "  1%|▎                                        | 12/1415 [00:06<12:55,  1.81it/s]\u001b[A\n",
      "  1%|▍                                        | 13/1415 [00:07<12:49,  1.82it/s]\u001b[A\n",
      "  1%|▍                                        | 14/1415 [00:07<12:32,  1.86it/s]\u001b[A\n",
      "  1%|▍                                        | 15/1415 [00:08<12:58,  1.80it/s]\u001b[A\n",
      "  1%|▍                                        | 16/1415 [00:08<12:56,  1.80it/s]\u001b[A\n",
      "  1%|▍                                        | 17/1415 [00:09<13:00,  1.79it/s]\u001b[A\n",
      "  1%|▌                                        | 18/1415 [00:10<13:09,  1.77it/s]\u001b[A\n",
      "  1%|▌                                        | 19/1415 [00:10<13:09,  1.77it/s]\u001b[A\n",
      "  1%|▌                                        | 20/1415 [00:11<12:48,  1.81it/s]\u001b[A\n",
      "  1%|▌                                        | 21/1415 [00:11<12:49,  1.81it/s]\u001b[A\n",
      "  2%|▋                                        | 22/1415 [00:12<12:32,  1.85it/s]\u001b[A\n",
      "  2%|▋                                        | 23/1415 [00:12<12:34,  1.84it/s]\u001b[A\n",
      "  2%|▋                                        | 24/1415 [00:13<12:24,  1.87it/s]\u001b[A\n",
      "  2%|▋                                        | 25/1415 [00:13<12:31,  1.85it/s]\u001b[A\n",
      "  2%|▊                                        | 26/1415 [00:14<12:35,  1.84it/s]\u001b[A\n",
      "  2%|▊                                        | 27/1415 [00:14<12:23,  1.87it/s]\u001b[A\n",
      "  2%|▊                                        | 28/1415 [00:15<12:21,  1.87it/s]\u001b[A\n",
      "  2%|▊                                        | 29/1415 [00:15<12:26,  1.86it/s]\u001b[A\n",
      "  2%|▊                                        | 30/1415 [00:16<12:33,  1.84it/s]\u001b[A\n",
      "  2%|▉                                        | 31/1415 [00:17<12:27,  1.85it/s]\u001b[A\n",
      "  2%|▉                                        | 32/1415 [00:17<12:29,  1.84it/s]\u001b[A\n",
      "  2%|▉                                        | 33/1415 [00:18<12:17,  1.87it/s]\u001b[A\n",
      "  2%|▉                                        | 34/1415 [00:18<12:25,  1.85it/s]\u001b[A\n",
      "  2%|█                                        | 35/1415 [00:19<12:16,  1.87it/s]\u001b[A\n",
      "  3%|█                                        | 36/1415 [00:19<12:18,  1.87it/s]\u001b[A\n",
      "  3%|█                                        | 37/1415 [00:20<12:09,  1.89it/s]\u001b[A\n",
      "  3%|█                                        | 38/1415 [00:20<12:23,  1.85it/s]\u001b[A\n",
      "  3%|█▏                                       | 39/1415 [00:21<12:14,  1.87it/s]\u001b[A\n",
      "  3%|█▏                                       | 40/1415 [00:21<12:31,  1.83it/s]\u001b[A\n",
      "  3%|█▏                                       | 41/1415 [00:22<12:17,  1.86it/s]\u001b[A\n",
      "  3%|█▏                                       | 42/1415 [00:23<12:29,  1.83it/s]\u001b[A\n",
      "  3%|█▏                                       | 43/1415 [00:23<12:57,  1.77it/s]\u001b[A\n",
      "  3%|█▎                                       | 44/1415 [00:24<12:43,  1.79it/s]\u001b[A\n",
      "  3%|█▎                                       | 45/1415 [00:24<12:42,  1.80it/s]\u001b[A\n",
      "  3%|█▎                                       | 46/1415 [00:25<12:36,  1.81it/s]\u001b[A\n",
      "  3%|█▎                                       | 47/1415 [00:25<12:39,  1.80it/s]\u001b[A\n",
      "  3%|█▍                                       | 48/1415 [00:26<12:22,  1.84it/s]\u001b[A\n",
      "  3%|█▍                                       | 49/1415 [00:26<12:23,  1.84it/s]\u001b[A\n",
      "  4%|█▍                                       | 50/1415 [00:27<12:32,  1.81it/s]\u001b[A\n",
      "  4%|█▍                                       | 51/1415 [00:28<12:48,  1.77it/s]\u001b[A\n",
      "  4%|█▌                                       | 52/1415 [00:28<12:36,  1.80it/s]\u001b[A\n",
      "  4%|█▌                                       | 53/1415 [00:29<12:46,  1.78it/s]\u001b[A\n",
      "  4%|█▌                                       | 54/1415 [00:29<12:40,  1.79it/s]\u001b[A\n",
      "  4%|█▌                                       | 55/1415 [00:30<12:54,  1.76it/s]\u001b[A\n",
      "  4%|█▌                                       | 56/1415 [00:30<12:43,  1.78it/s]\u001b[A\n",
      "  4%|█▋                                       | 57/1415 [00:31<12:43,  1.78it/s]\u001b[A\n",
      "  4%|█▋                                       | 58/1415 [00:31<12:33,  1.80it/s]\u001b[A\n",
      "  4%|█▋                                       | 59/1415 [00:32<12:38,  1.79it/s]\u001b[A\n",
      "  4%|█▋                                       | 60/1415 [00:33<12:37,  1.79it/s]\u001b[A\n",
      "  4%|█▊                                       | 61/1415 [00:33<12:35,  1.79it/s]\u001b[A\n",
      "  4%|█▊                                       | 62/1415 [00:34<12:25,  1.81it/s]\u001b[A\n",
      "  4%|█▊                                       | 63/1415 [00:34<12:35,  1.79it/s]\u001b[A\n",
      "  5%|█▊                                       | 64/1415 [00:35<12:50,  1.75it/s]\u001b[A\n",
      "  5%|█▉                                       | 65/1415 [00:35<12:51,  1.75it/s]\u001b[A\n",
      "  5%|█▉                                       | 66/1415 [00:36<13:03,  1.72it/s]\u001b[A\n",
      "  5%|█▉                                       | 67/1415 [00:37<12:59,  1.73it/s]\u001b[A\n",
      "  5%|█▉                                       | 68/1415 [00:37<13:07,  1.71it/s]\u001b[A\n",
      "  5%|█▉                                       | 69/1415 [00:38<12:53,  1.74it/s]\u001b[A\n",
      "  5%|██                                       | 70/1415 [00:38<12:58,  1.73it/s]\u001b[A\n",
      "  5%|██                                       | 71/1415 [00:39<12:45,  1.76it/s]\u001b[A\n",
      "  5%|██                                       | 72/1415 [00:40<13:15,  1.69it/s]\u001b[A\n",
      "  5%|██                                       | 73/1415 [00:40<13:15,  1.69it/s]\u001b[A\n",
      "  5%|██▏                                      | 74/1415 [00:41<13:14,  1.69it/s]\u001b[A\n",
      "  5%|██▏                                      | 75/1415 [00:41<13:04,  1.71it/s]\u001b[A\n",
      "  5%|██▏                                      | 76/1415 [00:42<13:08,  1.70it/s]\u001b[A\n",
      "  5%|██▏                                      | 77/1415 [00:42<13:12,  1.69it/s]\u001b[A\n",
      "  6%|██▎                                      | 78/1415 [00:43<13:04,  1.70it/s]\u001b[A\n",
      "  6%|██▎                                      | 79/1415 [00:44<12:59,  1.71it/s]\u001b[A\n",
      "  6%|██▎                                      | 80/1415 [00:44<13:07,  1.70it/s]\u001b[A\n",
      "  6%|██▎                                      | 81/1415 [00:45<13:16,  1.68it/s]\u001b[A\n",
      "  6%|██▍                                      | 82/1415 [00:45<13:18,  1.67it/s]\u001b[A\n",
      "  6%|██▍                                      | 83/1415 [00:46<13:29,  1.65it/s]\u001b[A\n",
      "  6%|██▍                                      | 84/1415 [00:47<13:28,  1.65it/s]\u001b[A\n",
      "  6%|██▍                                      | 85/1415 [00:47<13:36,  1.63it/s]\u001b[A\n",
      "  6%|██▍                                      | 86/1415 [00:48<13:19,  1.66it/s]\u001b[A\n",
      "  6%|██▌                                      | 87/1415 [00:49<13:28,  1.64it/s]\u001b[A\n",
      "  6%|██▌                                      | 88/1415 [00:49<13:28,  1.64it/s]\u001b[A\n",
      "  6%|██▌                                      | 89/1415 [00:50<13:23,  1.65it/s]\u001b[A\n",
      "  6%|██▌                                      | 90/1415 [00:50<13:17,  1.66it/s]\u001b[A\n",
      "  6%|██▋                                      | 91/1415 [00:51<13:04,  1.69it/s]\u001b[A\n",
      "  7%|██▋                                      | 92/1415 [00:51<13:12,  1.67it/s]\u001b[A\n",
      "  7%|██▋                                      | 93/1415 [00:52<13:02,  1.69it/s]\u001b[A\n",
      "  7%|██▋                                      | 94/1415 [00:53<13:12,  1.67it/s]\u001b[A\n",
      "  7%|██▊                                      | 95/1415 [00:53<12:53,  1.71it/s]\u001b[A\n",
      "  7%|██▊                                      | 96/1415 [00:54<12:56,  1.70it/s]\u001b[A\n",
      "  7%|██▊                                      | 97/1415 [00:54<13:17,  1.65it/s]\u001b[A\n",
      "  7%|██▊                                      | 98/1415 [00:55<13:28,  1.63it/s]\u001b[A\n",
      "  7%|██▊                                      | 99/1415 [00:56<13:10,  1.67it/s]\u001b[A\n",
      "  7%|██▊                                     | 100/1415 [00:56<13:10,  1.66it/s]\u001b[A\n",
      "  7%|██▊                                     | 101/1415 [00:57<13:17,  1.65it/s]\u001b[A\n",
      "  7%|██▉                                     | 102/1415 [00:57<13:11,  1.66it/s]\u001b[A\n",
      "  7%|██▉                                     | 103/1415 [00:58<12:52,  1.70it/s]\u001b[A\n",
      "  7%|██▉                                     | 104/1415 [00:59<12:54,  1.69it/s]\u001b[A\n",
      "  7%|██▉                                     | 105/1415 [00:59<12:39,  1.72it/s]\u001b[A\n",
      "  7%|██▉                                     | 106/1415 [01:00<12:41,  1.72it/s]\u001b[A\n",
      "  8%|███                                     | 107/1415 [01:00<12:35,  1.73it/s]\u001b[A\n",
      "  8%|███                                     | 108/1415 [01:01<12:50,  1.70it/s]\u001b[A\n",
      "  8%|███                                     | 109/1415 [01:02<12:52,  1.69it/s]\u001b[A\n",
      "  8%|███                                     | 110/1415 [01:02<13:01,  1.67it/s]\u001b[A\n",
      "  8%|███▏                                    | 111/1415 [01:03<13:07,  1.66it/s]\u001b[A\n",
      "  8%|███▏                                    | 112/1415 [01:03<13:12,  1.64it/s]\u001b[A\n",
      "  8%|███▏                                    | 113/1415 [01:04<13:00,  1.67it/s]\u001b[A\n",
      "  8%|███▏                                    | 114/1415 [01:05<13:04,  1.66it/s]\u001b[A\n",
      "  8%|███▎                                    | 115/1415 [01:05<13:31,  1.60it/s]\u001b[A\n",
      "  8%|███▎                                    | 116/1415 [01:06<13:14,  1.64it/s]\u001b[A\n",
      "  8%|███▎                                    | 117/1415 [01:07<13:31,  1.60it/s]\u001b[A\n",
      "  8%|███▎                                    | 118/1415 [01:07<13:00,  1.66it/s]\u001b[A\n",
      "  8%|███▎                                    | 119/1415 [01:08<13:08,  1.64it/s]\u001b[A\n",
      "  8%|███▍                                    | 120/1415 [01:08<12:57,  1.66it/s]\u001b[A\n",
      "  9%|███▍                                    | 121/1415 [01:09<13:05,  1.65it/s]\u001b[A\n",
      "  9%|███▍                                    | 122/1415 [01:09<12:42,  1.70it/s]\u001b[A\n",
      "  9%|███▍                                    | 123/1415 [01:10<13:10,  1.64it/s]\u001b[A\n",
      "  9%|███▌                                    | 124/1415 [01:11<12:45,  1.69it/s]\u001b[A\n",
      "  9%|███▌                                    | 125/1415 [01:11<12:57,  1.66it/s]\u001b[A\n",
      "  9%|███▌                                    | 126/1415 [01:12<12:36,  1.70it/s]\u001b[A\n",
      "  9%|███▌                                    | 127/1415 [01:12<12:46,  1.68it/s]\u001b[A\n",
      "  9%|███▌                                    | 128/1415 [01:13<12:56,  1.66it/s]\u001b[A\n",
      "  9%|███▋                                    | 129/1415 [01:14<12:51,  1.67it/s]\u001b[A\n",
      "  9%|███▋                                    | 130/1415 [01:14<12:55,  1.66it/s]\u001b[A\n",
      "  9%|███▋                                    | 131/1415 [01:15<13:03,  1.64it/s]\u001b[A\n",
      "  9%|███▋                                    | 132/1415 [01:16<13:12,  1.62it/s]\u001b[A\n",
      "  9%|███▊                                    | 133/1415 [01:16<13:13,  1.61it/s]\u001b[A\n",
      "  9%|███▊                                    | 134/1415 [01:17<13:33,  1.57it/s]\u001b[A\n",
      " 10%|███▊                                    | 135/1415 [01:17<13:17,  1.60it/s]\u001b[A\n",
      " 10%|███▊                                    | 136/1415 [01:18<13:24,  1.59it/s]\u001b[A\n",
      " 10%|███▊                                    | 137/1415 [01:19<13:29,  1.58it/s]\u001b[A\n",
      " 10%|███▉                                    | 138/1415 [01:19<13:43,  1.55it/s]\u001b[A\n",
      " 10%|███▉                                    | 139/1415 [01:20<13:43,  1.55it/s]\u001b[A\n",
      " 10%|███▉                                    | 140/1415 [01:21<13:20,  1.59it/s]\u001b[A\n",
      " 10%|███▉                                    | 141/1415 [01:21<13:28,  1.58it/s]\u001b[A\n",
      " 10%|████                                    | 142/1415 [01:22<13:28,  1.57it/s]\u001b[A\n",
      " 10%|████                                    | 143/1415 [01:23<13:30,  1.57it/s]\u001b[A\n",
      " 10%|████                                    | 144/1415 [01:23<13:37,  1.55it/s]\u001b[A\n",
      " 10%|████                                    | 145/1415 [01:24<13:35,  1.56it/s]\u001b[A\n",
      " 10%|████▏                                   | 146/1415 [01:24<13:16,  1.59it/s]\u001b[A\n",
      " 10%|████▏                                   | 147/1415 [01:25<13:04,  1.62it/s]\u001b[A\n",
      " 10%|████▏                                   | 148/1415 [01:26<13:10,  1.60it/s]\u001b[A\n",
      " 11%|████▏                                   | 149/1415 [01:26<13:14,  1.59it/s]\u001b[A\n",
      " 11%|████▏                                   | 150/1415 [01:27<13:01,  1.62it/s]\u001b[A\n",
      " 11%|████▎                                   | 151/1415 [01:28<13:21,  1.58it/s]\u001b[A\n",
      " 11%|████▎                                   | 152/1415 [01:28<13:19,  1.58it/s]\u001b[A\n",
      " 11%|████▎                                   | 153/1415 [01:29<13:16,  1.58it/s]\u001b[A\n",
      " 11%|████▎                                   | 154/1415 [01:29<13:00,  1.62it/s]\u001b[A\n",
      " 11%|████▍                                   | 155/1415 [01:30<13:08,  1.60it/s]\u001b[A\n",
      " 11%|████▍                                   | 156/1415 [01:31<12:57,  1.62it/s]\u001b[A\n",
      " 11%|████▍                                   | 157/1415 [01:31<13:13,  1.59it/s]\u001b[A\n",
      " 11%|████▍                                   | 158/1415 [01:32<12:46,  1.64it/s]\u001b[A\n",
      " 11%|████▍                                   | 159/1415 [01:33<12:51,  1.63it/s]\u001b[A\n",
      " 11%|████▌                                   | 160/1415 [01:33<12:43,  1.64it/s]\u001b[A\n",
      " 11%|████▌                                   | 161/1415 [01:34<12:44,  1.64it/s]\u001b[A\n",
      " 11%|████▌                                   | 162/1415 [01:34<13:04,  1.60it/s]\u001b[A\n",
      " 12%|████▌                                   | 163/1415 [01:35<13:10,  1.58it/s]\u001b[A\n",
      " 12%|████▋                                   | 164/1415 [01:36<13:12,  1.58it/s]\u001b[A\n",
      " 12%|████▋                                   | 165/1415 [01:36<13:04,  1.59it/s]\u001b[A\n",
      " 12%|████▋                                   | 166/1415 [01:37<13:03,  1.59it/s]\u001b[A\n",
      " 12%|████▋                                   | 167/1415 [01:38<12:50,  1.62it/s]\u001b[A\n",
      " 12%|████▋                                   | 168/1415 [01:38<12:51,  1.62it/s]\u001b[A\n",
      " 12%|████▊                                   | 169/1415 [01:39<12:44,  1.63it/s]\u001b[A\n",
      " 12%|████▊                                   | 170/1415 [01:39<12:46,  1.62it/s]\u001b[A\n",
      " 12%|████▊                                   | 171/1415 [01:40<12:36,  1.64it/s]\u001b[A\n",
      " 12%|████▊                                   | 172/1415 [01:41<12:40,  1.63it/s]\u001b[A\n",
      " 12%|████▉                                   | 173/1415 [01:41<12:35,  1.64it/s]\u001b[A\n",
      " 12%|████▉                                   | 174/1415 [01:42<12:42,  1.63it/s]\u001b[A\n",
      " 12%|████▉                                   | 175/1415 [01:42<12:17,  1.68it/s]\u001b[A\n",
      " 12%|████▉                                   | 176/1415 [01:43<12:30,  1.65it/s]\u001b[A\n",
      " 13%|█████                                   | 177/1415 [01:44<12:25,  1.66it/s]\u001b[A\n",
      " 13%|█████                                   | 178/1415 [01:44<12:32,  1.64it/s]\u001b[A\n",
      " 13%|█████                                   | 179/1415 [01:45<12:35,  1.64it/s]\u001b[A\n",
      " 13%|█████                                   | 180/1415 [01:45<12:27,  1.65it/s]\u001b[A\n",
      " 13%|█████                                   | 181/1415 [01:46<12:35,  1.63it/s]\u001b[A\n",
      " 13%|█████▏                                  | 182/1415 [01:47<12:27,  1.65it/s]\u001b[A\n",
      " 13%|█████▏                                  | 183/1415 [01:47<12:33,  1.63it/s]\u001b[A\n",
      " 13%|█████▏                                  | 184/1415 [01:48<12:25,  1.65it/s]\u001b[A\n",
      " 13%|█████▏                                  | 185/1415 [01:48<12:33,  1.63it/s]\u001b[A\n",
      " 13%|█████▎                                  | 186/1415 [01:49<12:24,  1.65it/s]\u001b[A\n",
      " 13%|█████▎                                  | 187/1415 [01:50<12:15,  1.67it/s]\u001b[A\n",
      " 13%|█████▎                                  | 188/1415 [01:50<12:25,  1.65it/s]\u001b[A\n",
      " 13%|█████▎                                  | 189/1415 [01:51<12:19,  1.66it/s]\u001b[A\n",
      " 13%|█████▎                                  | 190/1415 [01:51<12:27,  1.64it/s]\u001b[A\n",
      " 13%|█████▍                                  | 191/1415 [01:52<12:21,  1.65it/s]\u001b[A\n",
      " 14%|█████▍                                  | 192/1415 [01:53<12:22,  1.65it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 14%|█████▍                                  | 193/1415 [01:53<12:18,  1.66it/s]\u001b[A\n",
      " 14%|█████▍                                  | 194/1415 [01:54<12:26,  1.64it/s]\u001b[A\n",
      " 14%|█████▌                                  | 195/1415 [01:55<12:25,  1.64it/s]\u001b[A\n",
      " 14%|█████▌                                  | 196/1415 [01:55<12:38,  1.61it/s]\u001b[A\n",
      " 14%|█████▌                                  | 197/1415 [01:56<12:37,  1.61it/s]\u001b[A\n",
      " 14%|█████▌                                  | 198/1415 [01:56<12:22,  1.64it/s]\u001b[A\n",
      " 14%|█████▋                                  | 199/1415 [01:57<12:36,  1.61it/s]\u001b[A\n",
      " 14%|█████▋                                  | 200/1415 [01:58<12:40,  1.60it/s]\u001b[A\n",
      " 14%|█████▋                                  | 201/1415 [01:58<12:42,  1.59it/s]\u001b[A\n",
      " 14%|█████▋                                  | 202/1415 [01:59<12:26,  1.62it/s]\u001b[A\n",
      " 14%|█████▋                                  | 203/1415 [01:59<12:27,  1.62it/s]\u001b[A\n",
      " 14%|█████▊                                  | 204/1415 [02:00<12:16,  1.64it/s]\u001b[A\n",
      " 14%|█████▊                                  | 205/1415 [02:01<12:20,  1.63it/s]\u001b[A\n",
      " 15%|█████▊                                  | 206/1415 [02:01<12:33,  1.60it/s]\u001b[A\n",
      " 15%|█████▊                                  | 207/1415 [02:02<12:24,  1.62it/s]\u001b[A\n",
      " 15%|█████▉                                  | 208/1415 [02:03<12:31,  1.61it/s]\u001b[A\n",
      " 15%|█████▉                                  | 209/1415 [02:03<12:19,  1.63it/s]\u001b[A\n",
      " 15%|█████▉                                  | 210/1415 [02:04<12:38,  1.59it/s]\u001b[A\n",
      " 15%|█████▉                                  | 211/1415 [02:04<12:38,  1.59it/s]\u001b[A\n",
      " 15%|█████▉                                  | 212/1415 [02:05<12:33,  1.60it/s]\u001b[A\n",
      " 15%|██████                                  | 213/1415 [02:06<12:17,  1.63it/s]\u001b[A\n",
      " 15%|██████                                  | 214/1415 [02:06<12:24,  1.61it/s]\u001b[A\n",
      " 15%|██████                                  | 215/1415 [02:07<12:23,  1.61it/s]\u001b[A\n",
      " 15%|██████                                  | 216/1415 [02:08<12:13,  1.63it/s]\u001b[A\n",
      " 15%|██████▏                                 | 217/1415 [02:08<12:34,  1.59it/s]\u001b[A\n",
      " 15%|██████▏                                 | 218/1415 [02:09<12:09,  1.64it/s]\u001b[A\n",
      " 15%|██████▏                                 | 219/1415 [02:09<12:25,  1.61it/s]\u001b[A\n",
      " 16%|██████▏                                 | 220/1415 [02:10<12:23,  1.61it/s]\u001b[A\n",
      " 16%|██████▏                                 | 221/1415 [02:11<12:29,  1.59it/s]\u001b[A\n",
      " 16%|██████▎                                 | 222/1415 [02:11<12:13,  1.63it/s]\u001b[A\n",
      " 16%|██████▎                                 | 223/1415 [02:12<12:16,  1.62it/s]\u001b[A\n",
      " 16%|██████▎                                 | 224/1415 [02:13<12:19,  1.61it/s]\u001b[A\n",
      " 16%|██████▎                                 | 225/1415 [02:13<12:06,  1.64it/s]\u001b[A\n",
      " 16%|██████▍                                 | 226/1415 [02:14<12:13,  1.62it/s]\u001b[A\n",
      " 16%|██████▍                                 | 227/1415 [02:14<12:03,  1.64it/s]\u001b[A\n",
      " 16%|██████▍                                 | 228/1415 [02:15<12:06,  1.63it/s]\u001b[A\n",
      " 16%|██████▍                                 | 229/1415 [02:16<11:48,  1.67it/s]\u001b[A\n",
      " 16%|██████▌                                 | 230/1415 [02:16<11:59,  1.65it/s]\u001b[A\n",
      " 16%|██████▌                                 | 231/1415 [02:17<11:51,  1.66it/s]\u001b[A\n",
      " 16%|██████▌                                 | 232/1415 [02:17<11:58,  1.65it/s]\u001b[A\n",
      " 16%|██████▌                                 | 233/1415 [02:18<12:02,  1.63it/s]\u001b[A\n",
      " 17%|██████▌                                 | 234/1415 [02:19<11:48,  1.67it/s]\u001b[A\n",
      " 17%|██████▋                                 | 235/1415 [02:19<11:59,  1.64it/s]\u001b[A\n",
      " 17%|██████▋                                 | 236/1415 [02:20<12:09,  1.62it/s]\u001b[A\n",
      " 17%|██████▋                                 | 237/1415 [02:20<12:14,  1.60it/s]\u001b[A\n",
      " 17%|██████▋                                 | 238/1415 [02:21<12:02,  1.63it/s]\u001b[A\n",
      " 17%|██████▊                                 | 239/1415 [02:22<12:02,  1.63it/s]\u001b[A\n",
      " 17%|██████▊                                 | 240/1415 [02:22<11:44,  1.67it/s]\u001b[A\n",
      " 17%|██████▊                                 | 241/1415 [02:23<11:51,  1.65it/s]\u001b[A\n",
      " 17%|██████▊                                 | 242/1415 [02:23<11:59,  1.63it/s]\u001b[A\n",
      " 17%|██████▊                                 | 243/1415 [02:24<11:56,  1.64it/s]\u001b[A\n",
      " 17%|██████▉                                 | 244/1415 [02:25<12:00,  1.63it/s]\u001b[A\n",
      " 17%|██████▉                                 | 245/1415 [02:25<11:55,  1.63it/s]\u001b[A\n",
      " 17%|██████▉                                 | 246/1415 [02:26<11:47,  1.65it/s]\u001b[A\n",
      " 17%|██████▉                                 | 247/1415 [02:27<11:50,  1.64it/s]\u001b[A\n",
      " 18%|███████                                 | 248/1415 [02:27<12:12,  1.59it/s]\u001b[A\n",
      " 18%|███████                                 | 249/1415 [02:28<12:02,  1.61it/s]\u001b[A\n",
      " 18%|███████                                 | 250/1415 [02:28<12:03,  1.61it/s]\u001b[A\n",
      " 18%|███████                                 | 251/1415 [02:29<12:05,  1.60it/s]\u001b[A\n",
      " 18%|███████                                 | 252/1415 [02:30<12:08,  1.60it/s]\u001b[A\n",
      " 18%|███████▏                                | 253/1415 [02:30<12:27,  1.55it/s]\u001b[A\n",
      " 18%|███████▏                                | 254/1415 [02:31<12:24,  1.56it/s]\u001b[A\n",
      " 18%|███████▏                                | 255/1415 [02:32<12:15,  1.58it/s]\u001b[A\n",
      " 18%|███████▏                                | 256/1415 [02:32<12:03,  1.60it/s]\u001b[A\n",
      " 18%|███████▎                                | 257/1415 [02:33<12:04,  1.60it/s]\u001b[A\n",
      " 18%|███████▎                                | 258/1415 [02:33<11:56,  1.62it/s]\u001b[A\n",
      " 18%|███████▎                                | 259/1415 [02:34<11:56,  1.61it/s]\u001b[A\n",
      " 18%|███████▎                                | 260/1415 [02:35<12:07,  1.59it/s]\u001b[A\n",
      " 18%|███████▍                                | 261/1415 [02:35<12:01,  1.60it/s]\u001b[A\n",
      " 19%|███████▍                                | 262/1415 [02:36<12:00,  1.60it/s]\u001b[A\n",
      " 19%|███████▍                                | 263/1415 [02:37<11:49,  1.62it/s]\u001b[A\n",
      " 19%|███████▍                                | 264/1415 [02:37<11:53,  1.61it/s]\u001b[A\n",
      " 19%|███████▍                                | 265/1415 [02:38<11:48,  1.62it/s]\u001b[A\n",
      " 19%|███████▌                                | 266/1415 [02:38<11:52,  1.61it/s]\u001b[A\n",
      " 19%|███████▌                                | 267/1415 [02:39<11:28,  1.67it/s]\u001b[A\n",
      " 19%|███████▌                                | 268/1415 [02:40<11:37,  1.64it/s]\u001b[A\n",
      " 19%|███████▌                                | 269/1415 [02:40<11:48,  1.62it/s]\u001b[A\n",
      " 19%|███████▋                                | 270/1415 [02:41<11:43,  1.63it/s]\u001b[A\n",
      " 19%|███████▋                                | 271/1415 [02:41<11:48,  1.62it/s]\u001b[A\n",
      " 19%|███████▋                                | 272/1415 [02:42<11:39,  1.63it/s]\u001b[A\n",
      " 19%|███████▋                                | 273/1415 [02:43<11:46,  1.62it/s]\u001b[A\n",
      " 19%|███████▋                                | 274/1415 [02:43<11:46,  1.61it/s]\u001b[A\n",
      " 19%|███████▊                                | 275/1415 [02:44<11:48,  1.61it/s]\u001b[A\n",
      " 20%|███████▊                                | 276/1415 [02:45<11:38,  1.63it/s]\u001b[A\n",
      " 20%|███████▊                                | 277/1415 [02:45<11:38,  1.63it/s]\u001b[A\n",
      " 20%|███████▊                                | 278/1415 [02:46<11:38,  1.63it/s]\u001b[A\n",
      " 20%|███████▉                                | 279/1415 [02:46<11:31,  1.64it/s]\u001b[A\n",
      " 20%|███████▉                                | 280/1415 [02:47<11:35,  1.63it/s]\u001b[A\n",
      " 20%|███████▉                                | 281/1415 [02:48<11:28,  1.65it/s]\u001b[A\n",
      " 20%|███████▉                                | 282/1415 [02:48<11:34,  1.63it/s]\u001b[A\n",
      " 20%|████████                                | 283/1415 [02:49<11:26,  1.65it/s]\u001b[A\n",
      " 20%|████████                                | 284/1415 [02:49<11:33,  1.63it/s]\u001b[A\n",
      " 20%|████████                                | 285/1415 [02:50<11:30,  1.64it/s]\u001b[A\n",
      " 20%|████████                                | 286/1415 [02:51<11:30,  1.63it/s]\u001b[A\n",
      " 20%|████████                                | 287/1415 [02:51<11:43,  1.60it/s]\u001b[A\n",
      " 20%|████████▏                               | 288/1415 [02:52<11:51,  1.59it/s]\u001b[A\n",
      " 20%|████████▏                               | 289/1415 [02:53<11:44,  1.60it/s]\u001b[A\n",
      " 20%|████████▏                               | 290/1415 [02:53<11:30,  1.63it/s]\u001b[A\n",
      " 21%|████████▏                               | 291/1415 [02:54<11:32,  1.62it/s]\u001b[A\n",
      " 21%|████████▎                               | 292/1415 [02:54<11:21,  1.65it/s]\u001b[A\n",
      " 21%|████████▎                               | 293/1415 [02:55<11:25,  1.64it/s]\u001b[A\n",
      " 21%|████████▎                               | 294/1415 [02:56<11:33,  1.62it/s]\u001b[A\n",
      " 21%|████████▎                               | 295/1415 [02:56<11:45,  1.59it/s]\u001b[A\n",
      " 21%|████████▎                               | 296/1415 [02:57<11:58,  1.56it/s]\u001b[A\n",
      " 21%|████████▍                               | 297/1415 [02:58<11:55,  1.56it/s]\u001b[A\n",
      " 21%|████████▍                               | 298/1415 [02:58<12:08,  1.53it/s]\u001b[A\n",
      " 21%|████████▍                               | 299/1415 [02:59<12:01,  1.55it/s]\u001b[A\n",
      " 21%|████████▍                               | 300/1415 [03:00<12:07,  1.53it/s]\u001b[A\n",
      " 21%|████████▌                               | 301/1415 [03:00<11:49,  1.57it/s]\u001b[A\n",
      " 21%|████████▌                               | 302/1415 [03:01<11:47,  1.57it/s]\u001b[A\n",
      " 21%|████████▌                               | 303/1415 [03:01<11:35,  1.60it/s]\u001b[A\n",
      " 21%|████████▌                               | 304/1415 [03:02<11:36,  1.60it/s]\u001b[A\n",
      " 22%|████████▌                               | 305/1415 [03:03<11:37,  1.59it/s]\u001b[A\n",
      " 22%|████████▋                               | 306/1415 [03:03<11:36,  1.59it/s]\u001b[A\n",
      " 22%|████████▋                               | 307/1415 [03:04<11:38,  1.59it/s]\u001b[A\n",
      " 22%|████████▋                               | 308/1415 [03:05<11:27,  1.61it/s]\u001b[A\n",
      " 22%|████████▋                               | 309/1415 [03:05<11:27,  1.61it/s]\u001b[A\n",
      " 22%|████████▊                               | 310/1415 [03:06<11:15,  1.64it/s]\u001b[A\n",
      " 22%|████████▊                               | 311/1415 [03:06<11:22,  1.62it/s]\u001b[A\n",
      " 22%|████████▊                               | 312/1415 [03:07<11:14,  1.64it/s]\u001b[A\n",
      " 22%|████████▊                               | 313/1415 [03:08<11:20,  1.62it/s]\u001b[A\n",
      " 22%|████████▉                               | 314/1415 [03:08<11:24,  1.61it/s]\u001b[A\n",
      " 22%|████████▉                               | 315/1415 [03:09<11:00,  1.67it/s]\u001b[A\n",
      " 22%|████████▉                               | 316/1415 [03:09<10:57,  1.67it/s]\u001b[A\n",
      " 22%|████████▉                               | 317/1415 [03:10<11:01,  1.66it/s]\u001b[A\n",
      " 22%|████████▉                               | 318/1415 [03:11<11:13,  1.63it/s]\u001b[A\n",
      " 23%|█████████                               | 319/1415 [03:11<11:07,  1.64it/s]\u001b[A\n",
      " 23%|█████████                               | 320/1415 [03:12<11:12,  1.63it/s]\u001b[A\n",
      " 23%|█████████                               | 321/1415 [03:12<11:15,  1.62it/s]\u001b[A\n",
      " 23%|█████████                               | 322/1415 [03:13<11:29,  1.59it/s]\u001b[A\n",
      " 23%|█████████▏                              | 323/1415 [03:14<11:24,  1.60it/s]\u001b[A\n",
      " 23%|█████████▏                              | 324/1415 [03:14<11:13,  1.62it/s]\u001b[A\n",
      " 23%|█████████▏                              | 325/1415 [03:15<11:15,  1.61it/s]\u001b[A\n",
      " 23%|█████████▏                              | 326/1415 [03:16<10:50,  1.68it/s]\u001b[A\n",
      " 23%|█████████▏                              | 327/1415 [03:16<10:58,  1.65it/s]\u001b[A\n",
      " 23%|█████████▎                              | 328/1415 [03:17<10:59,  1.65it/s]\u001b[A\n",
      " 23%|█████████▎                              | 329/1415 [03:17<11:03,  1.64it/s]\u001b[A\n",
      " 23%|█████████▎                              | 330/1415 [03:18<10:55,  1.66it/s]\u001b[A\n",
      " 23%|█████████▎                              | 331/1415 [03:19<11:13,  1.61it/s]\u001b[A\n",
      " 23%|█████████▍                              | 332/1415 [03:19<10:59,  1.64it/s]\u001b[A\n",
      " 24%|█████████▍                              | 333/1415 [03:20<10:52,  1.66it/s]\u001b[A\n",
      " 24%|█████████▍                              | 334/1415 [03:20<11:09,  1.62it/s]\u001b[A\n",
      " 24%|█████████▍                              | 335/1415 [03:21<11:13,  1.60it/s]\u001b[A\n",
      " 24%|█████████▍                              | 336/1415 [03:22<11:29,  1.56it/s]\u001b[A\n",
      " 24%|█████████▌                              | 337/1415 [03:22<11:31,  1.56it/s]\u001b[A\n",
      " 24%|█████████▌                              | 338/1415 [03:23<11:41,  1.54it/s]\u001b[A\n",
      " 24%|█████████▌                              | 339/1415 [03:24<11:21,  1.58it/s]\u001b[A\n",
      " 24%|█████████▌                              | 340/1415 [03:24<11:16,  1.59it/s]\u001b[A\n",
      " 24%|█████████▋                              | 341/1415 [03:25<11:14,  1.59it/s]\u001b[A\n",
      " 24%|█████████▋                              | 342/1415 [03:26<11:17,  1.58it/s]\u001b[A\n",
      " 24%|█████████▋                              | 343/1415 [03:26<11:17,  1.58it/s]\u001b[A\n",
      " 24%|█████████▋                              | 344/1415 [03:27<11:05,  1.61it/s]\u001b[A\n",
      " 24%|█████████▊                              | 345/1415 [03:27<11:08,  1.60it/s]\u001b[A\n",
      " 24%|█████████▊                              | 346/1415 [03:28<10:59,  1.62it/s]\u001b[A\n",
      " 25%|█████████▊                              | 347/1415 [03:29<10:54,  1.63it/s]\u001b[A\n",
      " 25%|█████████▊                              | 348/1415 [03:29<10:46,  1.65it/s]\u001b[A\n",
      " 25%|█████████▊                              | 349/1415 [03:30<10:51,  1.64it/s]\u001b[A\n",
      " 25%|█████████▉                              | 350/1415 [03:30<11:08,  1.59it/s]\u001b[A\n",
      " 25%|█████████▉                              | 351/1415 [03:31<11:05,  1.60it/s]\u001b[A\n",
      " 25%|█████████▉                              | 352/1415 [03:32<11:18,  1.57it/s]\u001b[A\n",
      " 25%|█████████▉                              | 353/1415 [03:32<11:05,  1.60it/s]\u001b[A\n",
      " 25%|██████████                              | 354/1415 [03:33<11:06,  1.59it/s]\u001b[A\n",
      " 25%|██████████                              | 355/1415 [03:34<11:01,  1.60it/s]\u001b[A\n",
      " 25%|██████████                              | 356/1415 [03:34<11:07,  1.59it/s]\u001b[A\n",
      " 25%|██████████                              | 357/1415 [03:35<10:46,  1.64it/s]\u001b[A\n",
      " 25%|██████████                              | 358/1415 [03:35<10:50,  1.62it/s]\u001b[A\n",
      " 25%|██████████▏                             | 359/1415 [03:36<10:56,  1.61it/s]\u001b[A\n",
      " 25%|██████████▏                             | 360/1415 [03:37<11:01,  1.59it/s]\u001b[A\n",
      " 26%|██████████▏                             | 361/1415 [03:37<11:03,  1.59it/s]\u001b[A\n",
      " 26%|██████████▏                             | 362/1415 [03:38<10:53,  1.61it/s]\u001b[A\n",
      " 26%|██████████▎                             | 363/1415 [03:39<10:43,  1.63it/s]\u001b[A\n",
      " 26%|██████████▎                             | 364/1415 [03:39<10:37,  1.65it/s]\u001b[A\n",
      " 26%|██████████▎                             | 365/1415 [03:40<10:33,  1.66it/s]\u001b[A\n",
      " 26%|██████████▎                             | 366/1415 [03:40<10:31,  1.66it/s]\u001b[A\n",
      " 26%|██████████▎                             | 367/1415 [03:41<10:50,  1.61it/s]\u001b[A\n",
      " 26%|██████████▍                             | 368/1415 [03:42<11:01,  1.58it/s]\u001b[A\n",
      " 26%|██████████▍                             | 369/1415 [03:42<10:57,  1.59it/s]\u001b[A\n",
      " 26%|██████████▍                             | 370/1415 [03:43<11:15,  1.55it/s]\u001b[A\n",
      " 26%|██████████▍                             | 371/1415 [03:44<11:03,  1.57it/s]\u001b[A\n",
      " 26%|██████████▌                             | 372/1415 [03:44<11:13,  1.55it/s]\u001b[A\n",
      " 26%|██████████▌                             | 373/1415 [03:45<11:00,  1.58it/s]\u001b[A\n",
      " 26%|██████████▌                             | 374/1415 [03:46<11:00,  1.58it/s]\u001b[A\n",
      " 27%|██████████▌                             | 375/1415 [03:46<10:48,  1.60it/s]\u001b[A\n",
      " 27%|██████████▋                             | 376/1415 [03:47<10:47,  1.61it/s]\u001b[A\n",
      " 27%|██████████▋                             | 377/1415 [03:47<10:48,  1.60it/s]\u001b[A\n",
      " 27%|██████████▋                             | 378/1415 [03:48<10:40,  1.62it/s]\u001b[A\n",
      " 27%|██████████▋                             | 379/1415 [03:49<10:58,  1.57it/s]\u001b[A\n",
      " 27%|██████████▋                             | 380/1415 [03:49<10:56,  1.58it/s]\u001b[A\n",
      " 27%|██████████▊                             | 381/1415 [03:50<10:56,  1.58it/s]\u001b[A\n",
      " 27%|██████████▊                             | 382/1415 [03:51<11:01,  1.56it/s]\u001b[A\n",
      " 27%|██████████▊                             | 383/1415 [03:51<11:12,  1.54it/s]\u001b[A\n",
      " 27%|██████████▊                             | 384/1415 [03:52<10:56,  1.57it/s]\u001b[A\n",
      " 27%|██████████▉                             | 385/1415 [03:52<10:54,  1.57it/s]\u001b[A\n",
      " 27%|██████████▉                             | 386/1415 [03:53<10:46,  1.59it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 27%|██████████▉                             | 387/1415 [03:54<10:39,  1.61it/s]\u001b[A\n",
      " 27%|██████████▉                             | 388/1415 [03:54<10:46,  1.59it/s]\u001b[A\n",
      " 27%|██████████▉                             | 389/1415 [03:55<10:24,  1.64it/s]\u001b[A\n",
      " 28%|███████████                             | 390/1415 [03:56<10:30,  1.62it/s]\u001b[A\n",
      " 28%|███████████                             | 391/1415 [03:56<10:28,  1.63it/s]\u001b[A\n",
      " 28%|███████████                             | 392/1415 [03:57<10:32,  1.62it/s]\u001b[A\n",
      " 28%|███████████                             | 393/1415 [03:57<10:32,  1.62it/s]\u001b[A\n",
      " 28%|███████████▏                            | 394/1415 [03:58<10:34,  1.61it/s]\u001b[A\n",
      " 28%|███████████▏                            | 395/1415 [03:59<10:38,  1.60it/s]\u001b[A\n",
      " 28%|███████████▏                            | 396/1415 [03:59<10:26,  1.63it/s]\u001b[A\n",
      " 28%|███████████▏                            | 397/1415 [04:00<10:21,  1.64it/s]\u001b[A\n",
      " 28%|███████████▎                            | 398/1415 [04:00<10:14,  1.65it/s]\u001b[A\n",
      " 28%|███████████▎                            | 399/1415 [04:01<10:13,  1.66it/s]\u001b[A\n",
      " 28%|███████████▎                            | 400/1415 [04:02<10:12,  1.66it/s]\u001b[A\n",
      " 28%|███████████▎                            | 401/1415 [04:02<10:22,  1.63it/s]\u001b[A\n",
      " 28%|███████████▎                            | 402/1415 [04:03<10:08,  1.66it/s]\u001b[A\n",
      " 28%|███████████▍                            | 403/1415 [04:03<10:06,  1.67it/s]\u001b[A\n",
      " 29%|███████████▍                            | 404/1415 [04:04<10:06,  1.67it/s]\u001b[A\n",
      " 29%|███████████▍                            | 405/1415 [04:05<10:19,  1.63it/s]\u001b[A\n",
      " 29%|███████████▍                            | 406/1415 [04:05<10:32,  1.60it/s]\u001b[A\n",
      " 29%|███████████▌                            | 407/1415 [04:06<10:21,  1.62it/s]\u001b[A\n",
      " 29%|███████████▌                            | 408/1415 [04:07<10:30,  1.60it/s]\u001b[A\n",
      " 29%|███████████▌                            | 409/1415 [04:07<10:26,  1.61it/s]\u001b[A\n",
      " 29%|███████████▌                            | 410/1415 [04:08<10:31,  1.59it/s]\u001b[A\n",
      " 29%|███████████▌                            | 411/1415 [04:08<10:22,  1.61it/s]\u001b[A\n",
      " 29%|███████████▋                            | 412/1415 [04:09<10:25,  1.60it/s]\u001b[A\n",
      " 29%|███████████▋                            | 413/1415 [04:10<10:42,  1.56it/s]\u001b[A\n",
      " 29%|███████████▋                            | 414/1415 [04:10<10:28,  1.59it/s]\u001b[A\n",
      " 29%|███████████▋                            | 415/1415 [04:11<10:38,  1.57it/s]\u001b[A\n",
      " 29%|███████████▊                            | 416/1415 [04:12<10:24,  1.60it/s]\u001b[A\n",
      " 29%|███████████▊                            | 417/1415 [04:12<10:25,  1.60it/s]\u001b[A\n",
      " 30%|███████████▊                            | 418/1415 [04:13<10:13,  1.63it/s]\u001b[A\n",
      " 30%|███████████▊                            | 419/1415 [04:13<10:18,  1.61it/s]\u001b[A\n",
      " 30%|███████████▊                            | 420/1415 [04:14<10:03,  1.65it/s]\u001b[A\n",
      " 30%|███████████▉                            | 421/1415 [04:15<10:07,  1.64it/s]\u001b[A\n",
      " 30%|███████████▉                            | 422/1415 [04:15<10:21,  1.60it/s]\u001b[A\n",
      " 30%|███████████▉                            | 423/1415 [04:16<10:26,  1.58it/s]\u001b[A\n",
      " 30%|███████████▉                            | 424/1415 [04:17<10:36,  1.56it/s]\u001b[A\n",
      " 30%|████████████                            | 425/1415 [04:17<10:26,  1.58it/s]\u001b[A\n",
      " 30%|████████████                            | 426/1415 [04:18<10:25,  1.58it/s]\u001b[A\n",
      " 30%|████████████                            | 427/1415 [04:19<10:26,  1.58it/s]\u001b[A\n",
      " 30%|████████████                            | 428/1415 [04:19<10:28,  1.57it/s]\u001b[A\n",
      " 30%|████████████▏                           | 429/1415 [04:20<10:14,  1.60it/s]\u001b[A\n",
      " 30%|████████████▏                           | 430/1415 [04:20<10:16,  1.60it/s]\u001b[A\n",
      " 30%|████████████▏                           | 431/1415 [04:21<10:16,  1.60it/s]\u001b[A\n",
      " 31%|████████████▏                           | 432/1415 [04:22<10:05,  1.62it/s]\u001b[A\n",
      " 31%|████████████▏                           | 433/1415 [04:22<10:00,  1.64it/s]\u001b[A\n",
      " 31%|████████████▎                           | 434/1415 [04:23<10:07,  1.62it/s]\u001b[A\n",
      " 31%|████████████▎                           | 435/1415 [04:23<10:11,  1.60it/s]\u001b[A\n",
      " 31%|████████████▎                           | 436/1415 [04:24<10:00,  1.63it/s]\u001b[A\n",
      " 31%|████████████▎                           | 437/1415 [04:25<10:03,  1.62it/s]\u001b[A\n",
      " 31%|████████████▍                           | 438/1415 [04:25<09:59,  1.63it/s]\u001b[A\n",
      " 31%|████████████▍                           | 439/1415 [04:26<10:00,  1.63it/s]\u001b[A\n",
      " 31%|████████████▍                           | 440/1415 [04:27<10:05,  1.61it/s]\u001b[A\n",
      " 31%|████████████▍                           | 441/1415 [04:27<09:59,  1.62it/s]\u001b[A\n",
      " 31%|████████████▍                           | 442/1415 [04:28<10:14,  1.58it/s]\u001b[A\n",
      " 31%|████████████▌                           | 443/1415 [04:28<10:07,  1.60it/s]\u001b[A\n",
      " 31%|████████████▌                           | 444/1415 [04:29<10:07,  1.60it/s]\u001b[A\n",
      " 31%|████████████▌                           | 445/1415 [04:30<09:55,  1.63it/s]\u001b[A\n",
      " 32%|████████████▌                           | 446/1415 [04:30<10:09,  1.59it/s]\u001b[A\n",
      " 32%|████████████▋                           | 447/1415 [04:31<10:10,  1.58it/s]\u001b[A\n",
      " 32%|████████████▋                           | 448/1415 [04:32<10:11,  1.58it/s]\u001b[A\n",
      " 32%|████████████▋                           | 449/1415 [04:32<10:22,  1.55it/s]\u001b[A\n",
      " 32%|████████████▋                           | 450/1415 [04:33<09:59,  1.61it/s]\u001b[A\n",
      " 32%|████████████▋                           | 451/1415 [04:33<10:11,  1.58it/s]\u001b[A\n",
      " 32%|████████████▊                           | 452/1415 [04:34<10:01,  1.60it/s]\u001b[A\n",
      " 32%|████████████▊                           | 453/1415 [04:35<10:02,  1.60it/s]\u001b[A\n",
      " 32%|████████████▊                           | 454/1415 [04:35<09:50,  1.63it/s]\u001b[A\n",
      " 32%|████████████▊                           | 455/1415 [04:36<09:54,  1.62it/s]\u001b[A\n",
      " 32%|████████████▉                           | 456/1415 [04:37<09:58,  1.60it/s]\u001b[A\n",
      " 32%|████████████▉                           | 457/1415 [04:37<10:01,  1.59it/s]\u001b[A\n",
      " 32%|████████████▉                           | 458/1415 [04:38<10:03,  1.58it/s]\u001b[A\n",
      " 32%|████████████▉                           | 459/1415 [04:38<10:07,  1.57it/s]\u001b[A\n",
      " 33%|█████████████                           | 460/1415 [04:39<10:14,  1.55it/s]\u001b[A\n",
      " 33%|█████████████                           | 461/1415 [04:40<10:05,  1.57it/s]\u001b[A\n",
      " 33%|█████████████                           | 462/1415 [04:40<10:09,  1.56it/s]\u001b[A\n",
      " 33%|█████████████                           | 463/1415 [04:41<09:54,  1.60it/s]\u001b[A\n",
      " 33%|█████████████                           | 464/1415 [04:42<09:57,  1.59it/s]\u001b[A\n",
      " 33%|█████████████▏                          | 465/1415 [04:42<09:50,  1.61it/s]\u001b[A\n",
      " 33%|█████████████▏                          | 466/1415 [04:43<09:52,  1.60it/s]\u001b[A\n",
      " 33%|█████████████▏                          | 467/1415 [04:44<10:08,  1.56it/s]\u001b[A\n",
      " 33%|█████████████▏                          | 468/1415 [04:44<10:07,  1.56it/s]\u001b[A\n",
      " 33%|█████████████▎                          | 469/1415 [04:45<10:19,  1.53it/s]\u001b[A\n",
      " 33%|█████████████▎                          | 470/1415 [04:46<10:15,  1.53it/s]\u001b[A\n",
      " 33%|█████████████▎                          | 471/1415 [04:46<10:25,  1.51it/s]\u001b[A\n",
      " 33%|█████████████▎                          | 472/1415 [04:47<10:17,  1.53it/s]\u001b[A\n",
      " 33%|█████████████▎                          | 473/1415 [04:48<10:15,  1.53it/s]\u001b[A\n",
      " 33%|█████████████▍                          | 474/1415 [04:48<10:01,  1.57it/s]\u001b[A\n",
      " 34%|█████████████▍                          | 475/1415 [04:49<09:56,  1.58it/s]\u001b[A\n",
      " 34%|█████████████▍                          | 476/1415 [04:49<10:02,  1.56it/s]\u001b[A\n",
      " 34%|█████████████▍                          | 477/1415 [04:50<10:05,  1.55it/s]\u001b[A\n",
      " 34%|█████████████▌                          | 478/1415 [04:51<10:02,  1.56it/s]\u001b[A\n",
      " 34%|█████████████▌                          | 479/1415 [04:51<09:52,  1.58it/s]\u001b[A\n",
      " 34%|█████████████▌                          | 480/1415 [04:52<10:05,  1.55it/s]\u001b[A\n",
      " 34%|█████████████▌                          | 481/1415 [04:53<09:56,  1.57it/s]\u001b[A\n",
      " 34%|█████████████▋                          | 482/1415 [04:53<09:47,  1.59it/s]\u001b[A\n",
      " 34%|█████████████▋                          | 483/1415 [04:54<09:41,  1.60it/s]\u001b[A\n",
      " 34%|█████████████▋                          | 484/1415 [04:54<09:48,  1.58it/s]\u001b[A\n",
      " 34%|█████████████▋                          | 485/1415 [04:55<09:40,  1.60it/s]\u001b[A\n",
      " 34%|█████████████▋                          | 486/1415 [04:56<09:36,  1.61it/s]\u001b[A\n",
      " 34%|█████████████▊                          | 487/1415 [04:56<09:53,  1.56it/s]\u001b[A\n",
      " 34%|█████████████▊                          | 488/1415 [04:57<09:46,  1.58it/s]\u001b[A\n",
      " 35%|█████████████▊                          | 489/1415 [04:58<10:01,  1.54it/s]\u001b[A\n",
      " 35%|█████████████▊                          | 490/1415 [04:58<09:53,  1.56it/s]\u001b[A\n",
      " 35%|█████████████▉                          | 491/1415 [04:59<09:54,  1.55it/s]\u001b[A\n",
      " 35%|█████████████▉                          | 492/1415 [05:00<09:57,  1.55it/s]\u001b[A\n",
      " 35%|█████████████▉                          | 493/1415 [05:00<09:57,  1.54it/s]\u001b[A\n",
      " 35%|█████████████▉                          | 494/1415 [05:01<09:57,  1.54it/s]\u001b[A\n",
      " 35%|█████████████▉                          | 495/1415 [05:02<09:58,  1.54it/s]\u001b[A\n",
      " 35%|██████████████                          | 496/1415 [05:02<10:09,  1.51it/s]\u001b[A\n",
      " 35%|██████████████                          | 497/1415 [05:03<10:10,  1.50it/s]\u001b[A\n",
      " 35%|██████████████                          | 498/1415 [05:04<10:07,  1.51it/s]\u001b[A\n",
      " 35%|██████████████                          | 499/1415 [05:04<09:54,  1.54it/s]\u001b[A\n",
      " 35%|██████████████▏                         | 500/1415 [05:05<10:06,  1.51it/s]\u001b[A\n",
      " 35%|██████████████▏                         | 501/1415 [05:06<10:09,  1.50it/s]\u001b[A\n",
      " 35%|██████████████▏                         | 502/1415 [05:06<10:06,  1.51it/s]\u001b[A\n",
      " 36%|██████████████▏                         | 503/1415 [05:07<10:03,  1.51it/s]\u001b[A\n",
      " 36%|██████████████▏                         | 504/1415 [05:07<09:53,  1.53it/s]\u001b[A\n",
      " 36%|██████████████▎                         | 505/1415 [05:08<09:53,  1.53it/s]\u001b[A\n",
      " 36%|██████████████▎                         | 506/1415 [05:09<09:43,  1.56it/s]\u001b[A\n",
      " 36%|██████████████▎                         | 507/1415 [05:09<09:54,  1.53it/s]\u001b[A\n",
      " 36%|██████████████▎                         | 508/1415 [05:10<09:56,  1.52it/s]\u001b[A\n",
      " 36%|██████████████▍                         | 509/1415 [05:11<10:08,  1.49it/s]\u001b[A\n",
      " 36%|██████████████▍                         | 510/1415 [05:11<09:57,  1.51it/s]\u001b[A\n",
      " 36%|██████████████▍                         | 511/1415 [05:12<09:58,  1.51it/s]\u001b[A\n",
      " 36%|██████████████▍                         | 512/1415 [05:13<09:58,  1.51it/s]\u001b[A\n",
      " 36%|██████████████▌                         | 513/1415 [05:13<09:51,  1.53it/s]\u001b[A\n",
      " 36%|██████████████▌                         | 514/1415 [05:14<09:51,  1.52it/s]\u001b[A\n",
      " 36%|██████████████▌                         | 515/1415 [05:15<09:41,  1.55it/s]\u001b[A\n",
      " 36%|██████████████▌                         | 516/1415 [05:15<09:32,  1.57it/s]\u001b[A\n",
      " 37%|██████████████▌                         | 517/1415 [05:16<09:28,  1.58it/s]\u001b[A\n",
      " 37%|██████████████▋                         | 518/1415 [05:17<09:39,  1.55it/s]\u001b[A\n",
      " 37%|██████████████▋                         | 519/1415 [05:17<09:39,  1.55it/s]\u001b[A\n",
      " 37%|██████████████▋                         | 520/1415 [05:18<09:31,  1.57it/s]\u001b[A\n",
      " 37%|██████████████▋                         | 521/1415 [05:19<09:38,  1.55it/s]\u001b[A\n",
      " 37%|██████████████▊                         | 522/1415 [05:19<09:35,  1.55it/s]\u001b[A\n",
      " 37%|██████████████▊                         | 523/1415 [05:20<09:27,  1.57it/s]\u001b[A\n",
      " 37%|██████████████▊                         | 524/1415 [05:20<09:35,  1.55it/s]\u001b[A\n",
      " 37%|██████████████▊                         | 525/1415 [05:21<09:48,  1.51it/s]\u001b[A\n",
      " 37%|██████████████▊                         | 526/1415 [05:22<09:39,  1.53it/s]\u001b[A\n",
      " 37%|██████████████▉                         | 527/1415 [05:23<09:52,  1.50it/s]\u001b[A\n",
      " 37%|██████████████▉                         | 528/1415 [05:23<09:43,  1.52it/s]\u001b[A\n",
      " 37%|██████████████▉                         | 529/1415 [05:24<09:46,  1.51it/s]\u001b[A\n",
      " 37%|██████████████▉                         | 530/1415 [05:24<09:48,  1.50it/s]\u001b[A\n",
      " 38%|███████████████                         | 531/1415 [05:25<09:47,  1.51it/s]\u001b[A\n",
      " 38%|███████████████                         | 532/1415 [05:26<09:46,  1.51it/s]\u001b[A\n",
      " 38%|███████████████                         | 533/1415 [05:26<09:42,  1.51it/s]\u001b[A\n",
      " 38%|███████████████                         | 534/1415 [05:27<09:41,  1.52it/s]\u001b[A\n",
      " 38%|███████████████                         | 535/1415 [05:28<09:29,  1.55it/s]\u001b[A\n",
      " 38%|███████████████▏                        | 536/1415 [05:28<09:37,  1.52it/s]\u001b[A\n",
      " 38%|███████████████▏                        | 537/1415 [05:29<09:27,  1.55it/s]\u001b[A\n",
      " 38%|███████████████▏                        | 538/1415 [05:30<09:30,  1.54it/s]\u001b[A\n",
      " 38%|███████████████▏                        | 539/1415 [05:30<09:36,  1.52it/s]\u001b[A\n",
      " 38%|███████████████▎                        | 540/1415 [05:31<09:25,  1.55it/s]\u001b[A\n",
      " 38%|███████████████▎                        | 541/1415 [05:32<09:25,  1.54it/s]\u001b[A\n",
      " 38%|███████████████▎                        | 542/1415 [05:32<09:26,  1.54it/s]\u001b[A\n",
      " 38%|███████████████▎                        | 543/1415 [05:33<09:28,  1.53it/s]\u001b[A\n",
      " 38%|███████████████▍                        | 544/1415 [05:34<09:18,  1.56it/s]\u001b[A\n",
      " 39%|███████████████▍                        | 545/1415 [05:34<09:29,  1.53it/s]\u001b[A\n",
      " 39%|███████████████▍                        | 546/1415 [05:35<09:20,  1.55it/s]\u001b[A\n",
      " 39%|███████████████▍                        | 547/1415 [05:36<09:26,  1.53it/s]\u001b[A\n",
      " 39%|███████████████▍                        | 548/1415 [05:36<09:27,  1.53it/s]\u001b[A\n",
      " 39%|███████████████▌                        | 549/1415 [05:37<09:29,  1.52it/s]\u001b[A\n",
      " 39%|███████████████▌                        | 550/1415 [05:38<09:37,  1.50it/s]\u001b[A\n",
      " 39%|███████████████▌                        | 551/1415 [05:38<09:28,  1.52it/s]\u001b[A\n",
      " 39%|███████████████▌                        | 552/1415 [05:39<09:19,  1.54it/s]\u001b[A\n",
      " 39%|███████████████▋                        | 553/1415 [05:39<09:08,  1.57it/s]\u001b[A\n",
      " 39%|███████████████▋                        | 554/1415 [05:40<09:24,  1.53it/s]\u001b[A\n",
      " 39%|███████████████▋                        | 555/1415 [05:41<09:14,  1.55it/s]\u001b[A\n",
      " 39%|███████████████▋                        | 556/1415 [05:41<09:26,  1.52it/s]\u001b[A\n",
      " 39%|███████████████▋                        | 557/1415 [05:42<09:33,  1.50it/s]\u001b[A\n",
      " 39%|███████████████▊                        | 558/1415 [05:43<09:27,  1.51it/s]\u001b[A\n",
      " 40%|███████████████▊                        | 559/1415 [05:43<09:33,  1.49it/s]\u001b[A\n",
      " 40%|███████████████▊                        | 560/1415 [05:44<09:30,  1.50it/s]\u001b[A\n",
      " 40%|███████████████▊                        | 561/1415 [05:45<09:27,  1.50it/s]\u001b[A\n",
      " 40%|███████████████▉                        | 562/1415 [05:45<09:24,  1.51it/s]\u001b[A\n",
      " 40%|███████████████▉                        | 563/1415 [05:46<09:21,  1.52it/s]\u001b[A\n",
      " 40%|███████████████▉                        | 564/1415 [05:47<09:23,  1.51it/s]\u001b[A\n",
      " 40%|███████████████▉                        | 565/1415 [05:47<09:28,  1.50it/s]\u001b[A\n",
      " 40%|████████████████                        | 566/1415 [05:48<09:32,  1.48it/s]\u001b[A\n",
      " 40%|████████████████                        | 567/1415 [05:49<09:22,  1.51it/s]\u001b[A\n",
      " 40%|████████████████                        | 568/1415 [05:49<09:19,  1.51it/s]\u001b[A\n",
      " 40%|████████████████                        | 569/1415 [05:50<09:15,  1.52it/s]\u001b[A\n",
      " 40%|████████████████                        | 570/1415 [05:51<09:16,  1.52it/s]\u001b[A\n",
      " 40%|████████████████▏                       | 571/1415 [05:51<09:06,  1.55it/s]\u001b[A\n",
      " 40%|████████████████▏                       | 572/1415 [05:52<09:03,  1.55it/s]\u001b[A\n",
      " 40%|████████████████▏                       | 573/1415 [05:53<08:53,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▏                       | 574/1415 [05:53<08:58,  1.56it/s]\u001b[A\n",
      " 41%|████████████████▎                       | 575/1415 [05:54<09:00,  1.55it/s]\u001b[A\n",
      " 41%|████████████████▎                       | 576/1415 [05:55<08:44,  1.60it/s]\u001b[A\n",
      " 41%|████████████████▎                       | 577/1415 [05:55<08:46,  1.59it/s]\u001b[A\n",
      " 41%|████████████████▎                       | 578/1415 [05:56<08:40,  1.61it/s]\u001b[A\n",
      " 41%|████████████████▎                       | 579/1415 [05:56<08:50,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▍                       | 580/1415 [05:57<08:44,  1.59it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 41%|████████████████▍                       | 581/1415 [05:58<08:47,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▍                       | 582/1415 [05:58<08:40,  1.60it/s]\u001b[A\n",
      " 41%|████████████████▍                       | 583/1415 [05:59<08:44,  1.59it/s]\u001b[A\n",
      " 41%|████████████████▌                       | 584/1415 [06:00<08:46,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▌                       | 585/1415 [06:00<08:44,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▌                       | 586/1415 [06:01<08:45,  1.58it/s]\u001b[A\n",
      " 41%|████████████████▌                       | 587/1415 [06:01<08:43,  1.58it/s]\u001b[A\n",
      " 42%|████████████████▌                       | 588/1415 [06:02<08:45,  1.57it/s]\u001b[A\n",
      " 42%|████████████████▋                       | 589/1415 [06:03<08:36,  1.60it/s]\u001b[A\n",
      " 42%|████████████████▋                       | 590/1415 [06:03<08:36,  1.60it/s]\u001b[A\n",
      " 42%|████████████████▋                       | 591/1415 [06:04<08:31,  1.61it/s]\u001b[A\n",
      " 42%|████████████████▋                       | 592/1415 [06:05<08:43,  1.57it/s]\u001b[A\n",
      " 42%|████████████████▊                       | 593/1415 [06:05<08:56,  1.53it/s]\u001b[A\n",
      " 42%|████████████████▊                       | 594/1415 [06:06<08:56,  1.53it/s]\u001b[A\n",
      " 42%|████████████████▊                       | 595/1415 [06:07<08:53,  1.54it/s]\u001b[A\n",
      " 42%|████████████████▊                       | 596/1415 [06:07<08:42,  1.57it/s]\u001b[A\n",
      " 42%|████████████████▉                       | 597/1415 [06:08<08:42,  1.57it/s]\u001b[A\n",
      " 42%|████████████████▉                       | 598/1415 [06:08<08:32,  1.60it/s]\u001b[A\n",
      " 42%|████████████████▉                       | 599/1415 [06:09<08:36,  1.58it/s]\u001b[A\n",
      " 42%|████████████████▉                       | 600/1415 [06:10<08:39,  1.57it/s]\u001b[A\n",
      " 42%|████████████████▉                       | 601/1415 [06:10<08:42,  1.56it/s]\u001b[A\n",
      " 43%|█████████████████                       | 602/1415 [06:11<08:43,  1.55it/s]\u001b[A\n",
      " 43%|█████████████████                       | 603/1415 [06:12<08:34,  1.58it/s]\u001b[A\n",
      " 43%|█████████████████                       | 604/1415 [06:12<08:37,  1.57it/s]\u001b[A\n",
      " 43%|█████████████████                       | 605/1415 [06:13<08:29,  1.59it/s]\u001b[A\n",
      " 43%|█████████████████▏                      | 606/1415 [06:14<08:35,  1.57it/s]\u001b[A\n",
      " 43%|█████████████████▏                      | 607/1415 [06:14<08:23,  1.61it/s]\u001b[A\n",
      " 43%|█████████████████▏                      | 608/1415 [06:15<08:26,  1.59it/s]\u001b[A\n",
      " 43%|█████████████████▏                      | 609/1415 [06:15<08:27,  1.59it/s]\u001b[A\n",
      " 43%|█████████████████▏                      | 610/1415 [06:16<08:37,  1.56it/s]\u001b[A\n",
      " 43%|█████████████████▎                      | 611/1415 [06:17<08:40,  1.54it/s]\u001b[A\n",
      " 43%|█████████████████▎                      | 612/1415 [06:17<08:38,  1.55it/s]\u001b[A\n",
      " 43%|█████████████████▎                      | 613/1415 [06:18<08:43,  1.53it/s]\u001b[A\n",
      " 43%|█████████████████▎                      | 614/1415 [06:19<08:44,  1.53it/s]\u001b[A\n",
      " 43%|█████████████████▍                      | 615/1415 [06:19<08:59,  1.48it/s]\u001b[A\n",
      " 44%|█████████████████▍                      | 616/1415 [06:20<08:59,  1.48it/s]\u001b[A\n",
      " 44%|█████████████████▍                      | 617/1415 [06:21<08:57,  1.49it/s]\u001b[A\n",
      " 44%|█████████████████▍                      | 618/1415 [06:21<08:54,  1.49it/s]\u001b[A\n",
      " 44%|█████████████████▍                      | 619/1415 [06:22<08:58,  1.48it/s]\u001b[A\n",
      " 44%|█████████████████▌                      | 620/1415 [06:23<09:06,  1.45it/s]\u001b[A\n",
      " 44%|█████████████████▌                      | 621/1415 [06:24<08:56,  1.48it/s]\u001b[A\n",
      " 44%|█████████████████▌                      | 622/1415 [06:24<08:54,  1.48it/s]\u001b[A\n",
      " 44%|█████████████████▌                      | 623/1415 [06:25<08:41,  1.52it/s]\u001b[A\n",
      " 44%|█████████████████▋                      | 624/1415 [06:25<08:41,  1.52it/s]\u001b[A\n",
      " 44%|█████████████████▋                      | 625/1415 [06:26<08:37,  1.53it/s]\u001b[A\n",
      " 44%|█████████████████▋                      | 626/1415 [06:27<08:48,  1.49it/s]\u001b[A\n",
      " 44%|█████████████████▋                      | 627/1415 [06:27<08:42,  1.51it/s]\u001b[A\n",
      " 44%|█████████████████▊                      | 628/1415 [06:28<08:40,  1.51it/s]\u001b[A\n",
      " 44%|█████████████████▊                      | 629/1415 [06:29<08:49,  1.48it/s]\u001b[A\n",
      " 45%|█████████████████▊                      | 630/1415 [06:30<08:51,  1.48it/s]\u001b[A\n",
      " 45%|█████████████████▊                      | 631/1415 [06:30<09:01,  1.45it/s]\u001b[A\n",
      " 45%|█████████████████▊                      | 632/1415 [06:31<08:57,  1.46it/s]\u001b[A\n",
      " 45%|█████████████████▉                      | 633/1415 [06:32<08:56,  1.46it/s]\u001b[A\n",
      " 45%|█████████████████▉                      | 634/1415 [06:32<08:46,  1.48it/s]\u001b[A\n",
      " 45%|█████████████████▉                      | 635/1415 [06:33<08:43,  1.49it/s]\u001b[A\n",
      " 45%|█████████████████▉                      | 636/1415 [06:34<08:45,  1.48it/s]\u001b[A\n",
      " 45%|██████████████████                      | 637/1415 [06:34<08:42,  1.49it/s]\u001b[A\n",
      " 45%|██████████████████                      | 638/1415 [06:35<08:42,  1.49it/s]\u001b[A\n",
      " 45%|██████████████████                      | 639/1415 [06:36<08:35,  1.50it/s]\u001b[A\n",
      " 45%|██████████████████                      | 640/1415 [06:36<08:36,  1.50it/s]\u001b[A\n",
      " 45%|██████████████████                      | 641/1415 [06:37<08:34,  1.50it/s]\u001b[A\n",
      " 45%|██████████████████▏                     | 642/1415 [06:38<08:36,  1.50it/s]\u001b[A\n",
      " 45%|██████████████████▏                     | 643/1415 [06:38<08:39,  1.49it/s]\u001b[A\n",
      " 46%|██████████████████▏                     | 644/1415 [06:39<08:56,  1.44it/s]\u001b[A\n",
      " 46%|██████████████████▏                     | 645/1415 [06:40<08:43,  1.47it/s]\u001b[A\n",
      " 46%|██████████████████▎                     | 646/1415 [06:40<08:45,  1.46it/s]\u001b[A\n",
      " 46%|██████████████████▎                     | 647/1415 [06:41<08:49,  1.45it/s]\u001b[A\n",
      " 46%|██████████████████▎                     | 648/1415 [06:42<08:36,  1.48it/s]\u001b[A\n",
      " 46%|██████████████████▎                     | 649/1415 [06:42<08:42,  1.47it/s]\u001b[A\n",
      " 46%|██████████████████▎                     | 650/1415 [06:43<08:33,  1.49it/s]\u001b[A\n",
      " 46%|██████████████████▍                     | 651/1415 [06:44<08:32,  1.49it/s]\u001b[A\n",
      " 46%|██████████████████▍                     | 652/1415 [06:44<08:34,  1.48it/s]\u001b[A\n",
      " 46%|██████████████████▍                     | 653/1415 [06:45<08:41,  1.46it/s]\u001b[A\n",
      " 46%|██████████████████▍                     | 654/1415 [06:46<08:41,  1.46it/s]\u001b[A\n",
      " 46%|██████████████████▌                     | 655/1415 [06:47<08:49,  1.44it/s]\u001b[A\n",
      " 46%|██████████████████▌                     | 656/1415 [06:47<08:56,  1.41it/s]\u001b[A\n",
      " 46%|██████████████████▌                     | 657/1415 [06:48<08:42,  1.45it/s]\u001b[A\n",
      " 47%|██████████████████▌                     | 658/1415 [06:49<08:39,  1.46it/s]\u001b[A\n",
      " 47%|██████████████████▋                     | 659/1415 [06:49<08:27,  1.49it/s]\u001b[A\n",
      " 47%|██████████████████▋                     | 660/1415 [06:50<08:27,  1.49it/s]\u001b[A\n",
      " 47%|██████████████████▋                     | 661/1415 [06:51<08:19,  1.51it/s]\u001b[A\n",
      " 47%|██████████████████▋                     | 662/1415 [06:51<08:23,  1.50it/s]\u001b[A\n",
      " 47%|██████████████████▋                     | 663/1415 [06:52<08:16,  1.52it/s]\u001b[A\n",
      " 47%|██████████████████▊                     | 664/1415 [06:53<08:28,  1.48it/s]\u001b[A\n",
      " 47%|██████████████████▊                     | 665/1415 [06:53<08:33,  1.46it/s]\u001b[A\n",
      " 47%|██████████████████▊                     | 666/1415 [06:54<08:26,  1.48it/s]\u001b[A\n",
      " 47%|██████████████████▊                     | 667/1415 [06:55<08:24,  1.48it/s]\u001b[A\n",
      " 47%|██████████████████▉                     | 668/1415 [06:55<08:17,  1.50it/s]\u001b[A\n",
      " 47%|██████████████████▉                     | 669/1415 [06:56<08:19,  1.49it/s]\u001b[A\n",
      " 47%|██████████████████▉                     | 670/1415 [06:57<08:17,  1.50it/s]\u001b[A\n",
      " 47%|██████████████████▉                     | 671/1415 [06:57<08:22,  1.48it/s]\u001b[A\n",
      " 47%|██████████████████▉                     | 672/1415 [06:58<08:17,  1.49it/s]\u001b[A\n",
      " 48%|███████████████████                     | 673/1415 [06:59<08:14,  1.50it/s]\u001b[A\n",
      " 48%|███████████████████                     | 674/1415 [06:59<08:13,  1.50it/s]\u001b[A\n",
      " 48%|███████████████████                     | 675/1415 [07:00<08:05,  1.52it/s]\u001b[A\n",
      " 48%|███████████████████                     | 676/1415 [07:01<08:14,  1.49it/s]\u001b[A\n",
      " 48%|███████████████████▏                    | 677/1415 [07:01<08:11,  1.50it/s]\u001b[A\n",
      " 48%|███████████████████▏                    | 678/1415 [07:02<08:20,  1.47it/s]\u001b[A\n",
      " 48%|███████████████████▏                    | 679/1415 [07:03<08:10,  1.50it/s]\u001b[A\n",
      " 48%|███████████████████▏                    | 680/1415 [07:03<08:15,  1.48it/s]\u001b[A\n",
      " 48%|███████████████████▎                    | 681/1415 [07:04<08:04,  1.52it/s]\u001b[A\n",
      " 48%|███████████████████▎                    | 682/1415 [07:05<08:05,  1.51it/s]\u001b[A\n",
      " 48%|███████████████████▎                    | 683/1415 [07:05<08:05,  1.51it/s]\u001b[A\n",
      " 48%|███████████████████▎                    | 684/1415 [07:06<08:07,  1.50it/s]\u001b[A\n",
      " 48%|███████████████████▎                    | 685/1415 [07:07<08:11,  1.48it/s]\u001b[A\n",
      " 48%|███████████████████▍                    | 686/1415 [07:07<08:00,  1.52it/s]\u001b[A\n",
      " 49%|███████████████████▍                    | 687/1415 [07:08<07:58,  1.52it/s]\u001b[A\n",
      " 49%|███████████████████▍                    | 688/1415 [07:09<07:52,  1.54it/s]\u001b[A\n",
      " 49%|███████████████████▍                    | 689/1415 [07:09<07:57,  1.52it/s]\u001b[A\n",
      " 49%|███████████████████▌                    | 690/1415 [07:10<08:01,  1.51it/s]\u001b[A\n",
      " 49%|███████████████████▌                    | 691/1415 [07:11<08:01,  1.50it/s]\u001b[A\n",
      " 49%|███████████████████▌                    | 692/1415 [07:11<08:02,  1.50it/s]\u001b[A\n",
      " 49%|███████████████████▌                    | 693/1415 [07:12<07:53,  1.52it/s]\u001b[A\n",
      " 49%|███████████████████▌                    | 694/1415 [07:12<07:51,  1.53it/s]\u001b[A\n",
      " 49%|███████████████████▋                    | 695/1415 [07:13<07:45,  1.55it/s]\u001b[A\n",
      " 49%|███████████████████▋                    | 696/1415 [07:14<07:57,  1.50it/s]\u001b[A\n",
      " 49%|███████████████████▋                    | 697/1415 [07:14<07:59,  1.50it/s]\u001b[A\n",
      " 49%|███████████████████▋                    | 698/1415 [07:15<08:06,  1.47it/s]\u001b[A\n",
      " 49%|███████████████████▊                    | 699/1415 [07:16<07:54,  1.51it/s]\u001b[A\n",
      " 49%|███████████████████▊                    | 700/1415 [07:17<07:59,  1.49it/s]\u001b[A\n",
      " 50%|███████████████████▊                    | 701/1415 [07:17<08:06,  1.47it/s]\u001b[A\n",
      " 50%|███████████████████▊                    | 702/1415 [07:18<07:54,  1.50it/s]\u001b[A\n",
      " 50%|███████████████████▊                    | 703/1415 [07:19<07:54,  1.50it/s]\u001b[A\n",
      " 50%|███████████████████▉                    | 704/1415 [07:19<07:54,  1.50it/s]\u001b[A\n",
      " 50%|███████████████████▉                    | 705/1415 [07:20<07:54,  1.50it/s]\u001b[A\n",
      " 50%|███████████████████▉                    | 706/1415 [07:20<07:48,  1.51it/s]\u001b[A\n",
      " 50%|███████████████████▉                    | 707/1415 [07:21<07:47,  1.52it/s]\u001b[A\n",
      " 50%|████████████████████                    | 708/1415 [07:22<07:38,  1.54it/s]\u001b[A\n",
      " 50%|████████████████████                    | 709/1415 [07:22<07:45,  1.52it/s]\u001b[A\n",
      " 50%|████████████████████                    | 710/1415 [07:23<07:44,  1.52it/s]\u001b[A\n",
      " 50%|████████████████████                    | 711/1415 [07:24<07:36,  1.54it/s]\u001b[A\n",
      " 50%|████████████████████▏                   | 712/1415 [07:24<07:45,  1.51it/s]\u001b[A\n",
      " 50%|████████████████████▏                   | 713/1415 [07:25<07:36,  1.54it/s]\u001b[A\n",
      " 50%|████████████████████▏                   | 714/1415 [07:26<07:37,  1.53it/s]\u001b[A\n",
      " 51%|████████████████████▏                   | 715/1415 [07:26<07:34,  1.54it/s]\u001b[A\n",
      " 51%|████████████████████▏                   | 716/1415 [07:27<07:37,  1.53it/s]\u001b[A\n",
      " 51%|████████████████████▎                   | 717/1415 [07:28<07:30,  1.55it/s]\u001b[A\n",
      " 51%|████████████████████▎                   | 718/1415 [07:28<07:34,  1.53it/s]\u001b[A\n",
      " 51%|████████████████████▎                   | 719/1415 [07:29<07:34,  1.53it/s]\u001b[A\n",
      " 51%|████████████████████▎                   | 720/1415 [07:30<07:26,  1.56it/s]\u001b[A\n",
      " 51%|████████████████████▍                   | 721/1415 [07:30<07:38,  1.51it/s]\u001b[A\n",
      " 51%|████████████████████▍                   | 722/1415 [07:31<07:31,  1.53it/s]\u001b[A\n",
      " 51%|████████████████████▍                   | 723/1415 [07:32<07:41,  1.50it/s]\u001b[A\n",
      " 51%|████████████████████▍                   | 724/1415 [07:32<07:37,  1.51it/s]\u001b[A\n",
      " 51%|████████████████████▍                   | 725/1415 [07:33<07:40,  1.50it/s]\u001b[A\n",
      " 51%|████████████████████▌                   | 726/1415 [07:34<07:49,  1.47it/s]\u001b[A\n",
      " 51%|████████████████████▌                   | 727/1415 [07:34<07:56,  1.44it/s]\u001b[A\n",
      " 51%|████████████████████▌                   | 728/1415 [07:35<08:00,  1.43it/s]\u001b[A\n",
      " 52%|████████████████████▌                   | 729/1415 [07:36<08:03,  1.42it/s]\u001b[A\n",
      " 52%|████████████████████▋                   | 730/1415 [07:37<08:09,  1.40it/s]\u001b[A\n",
      " 52%|████████████████████▋                   | 731/1415 [07:37<07:59,  1.43it/s]\u001b[A\n",
      " 52%|████████████████████▋                   | 732/1415 [07:38<07:59,  1.42it/s]\u001b[A\n",
      " 52%|████████████████████▋                   | 733/1415 [07:39<07:53,  1.44it/s]\u001b[A\n",
      " 52%|████████████████████▋                   | 734/1415 [07:39<07:58,  1.42it/s]\u001b[A\n",
      " 52%|████████████████████▊                   | 735/1415 [07:40<07:51,  1.44it/s]\u001b[A\n",
      " 52%|████████████████████▊                   | 736/1415 [07:41<08:02,  1.41it/s]\u001b[A\n",
      " 52%|████████████████████▊                   | 737/1415 [07:41<07:58,  1.42it/s]\u001b[A\n",
      " 52%|████████████████████▊                   | 738/1415 [07:42<07:52,  1.43it/s]\u001b[A\n",
      " 52%|████████████████████▉                   | 739/1415 [07:43<07:56,  1.42it/s]\u001b[A\n",
      " 52%|████████████████████▉                   | 740/1415 [07:44<07:47,  1.44it/s]\u001b[A\n",
      " 52%|████████████████████▉                   | 741/1415 [07:44<07:48,  1.44it/s]\u001b[A\n",
      " 52%|████████████████████▉                   | 742/1415 [07:45<07:49,  1.43it/s]\u001b[A\n",
      " 53%|█████████████████████                   | 743/1415 [07:46<07:47,  1.44it/s]\u001b[A\n",
      " 53%|█████████████████████                   | 744/1415 [07:46<07:41,  1.46it/s]\u001b[A\n",
      " 53%|█████████████████████                   | 745/1415 [07:47<07:42,  1.45it/s]\u001b[A\n",
      " 53%|█████████████████████                   | 746/1415 [07:48<07:48,  1.43it/s]\u001b[A\n",
      " 53%|█████████████████████                   | 747/1415 [07:48<07:36,  1.46it/s]\u001b[A\n",
      " 53%|█████████████████████▏                  | 748/1415 [07:49<07:35,  1.47it/s]\u001b[A\n",
      " 53%|█████████████████████▏                  | 749/1415 [07:50<07:28,  1.49it/s]\u001b[A\n",
      " 53%|█████████████████████▏                  | 750/1415 [07:50<07:33,  1.47it/s]\u001b[A\n",
      " 53%|█████████████████████▏                  | 751/1415 [07:51<07:29,  1.48it/s]\u001b[A\n",
      " 53%|█████████████████████▎                  | 752/1415 [07:52<07:30,  1.47it/s]\u001b[A\n",
      " 53%|█████████████████████▎                  | 753/1415 [07:52<07:28,  1.48it/s]\u001b[A\n",
      " 53%|█████████████████████▎                  | 754/1415 [07:53<07:29,  1.47it/s]\u001b[A\n",
      " 53%|█████████████████████▎                  | 755/1415 [07:54<07:38,  1.44it/s]\u001b[A\n",
      " 53%|█████████████████████▎                  | 756/1415 [07:55<07:40,  1.43it/s]\u001b[A\n",
      " 53%|█████████████████████▍                  | 757/1415 [07:55<07:40,  1.43it/s]\u001b[A\n",
      " 54%|█████████████████████▍                  | 758/1415 [07:56<07:28,  1.46it/s]\u001b[A\n",
      " 54%|█████████████████████▍                  | 759/1415 [07:57<07:35,  1.44it/s]\u001b[A\n",
      " 54%|█████████████████████▍                  | 760/1415 [07:57<07:38,  1.43it/s]\u001b[A\n",
      " 54%|█████████████████████▌                  | 761/1415 [07:58<07:45,  1.40it/s]\u001b[A\n",
      " 54%|█████████████████████▌                  | 762/1415 [07:59<07:42,  1.41it/s]\u001b[A\n",
      " 54%|█████████████████████▌                  | 763/1415 [07:59<07:43,  1.41it/s]\u001b[A\n",
      " 54%|█████████████████████▌                  | 764/1415 [08:00<07:49,  1.39it/s]\u001b[A\n",
      " 54%|█████████████████████▋                  | 765/1415 [08:01<07:39,  1.42it/s]\u001b[A\n",
      " 54%|█████████████████████▋                  | 766/1415 [08:02<07:38,  1.42it/s]\u001b[A\n",
      " 54%|█████████████████████▋                  | 767/1415 [08:02<07:32,  1.43it/s]\u001b[A\n",
      " 54%|█████████████████████▋                  | 768/1415 [08:03<07:49,  1.38it/s]\u001b[A\n",
      " 54%|█████████████████████▋                  | 769/1415 [08:04<07:42,  1.40it/s]\u001b[A\n",
      " 54%|█████████████████████▊                  | 770/1415 [08:04<07:47,  1.38it/s]\u001b[A\n",
      " 54%|█████████████████████▊                  | 771/1415 [08:05<07:33,  1.42it/s]\u001b[A\n",
      " 55%|█████████████████████▊                  | 772/1415 [08:06<07:38,  1.40it/s]\u001b[A\n",
      " 55%|█████████████████████▊                  | 773/1415 [08:07<07:42,  1.39it/s]\u001b[A\n",
      " 55%|█████████████████████▉                  | 774/1415 [08:07<07:39,  1.39it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 55%|█████████████████████▉                  | 775/1415 [08:08<07:42,  1.38it/s]\u001b[A\n",
      " 55%|█████████████████████▉                  | 776/1415 [08:09<07:28,  1.43it/s]\u001b[A\n",
      " 55%|█████████████████████▉                  | 777/1415 [08:09<07:25,  1.43it/s]\u001b[A\n",
      " 55%|█████████████████████▉                  | 778/1415 [08:10<07:24,  1.43it/s]\u001b[A\n",
      " 55%|██████████████████████                  | 779/1415 [08:11<07:21,  1.44it/s]\u001b[A\n",
      " 55%|██████████████████████                  | 780/1415 [08:11<07:18,  1.45it/s]\u001b[A\n",
      " 55%|██████████████████████                  | 781/1415 [08:12<07:18,  1.45it/s]\u001b[A\n",
      " 55%|██████████████████████                  | 782/1415 [08:13<07:19,  1.44it/s]\u001b[A\n",
      " 55%|██████████████████████▏                 | 783/1415 [08:14<07:13,  1.46it/s]\u001b[A\n",
      " 55%|██████████████████████▏                 | 784/1415 [08:14<07:23,  1.42it/s]\u001b[A\n",
      " 55%|██████████████████████▏                 | 785/1415 [08:15<07:15,  1.45it/s]\u001b[A\n",
      " 56%|██████████████████████▏                 | 786/1415 [08:16<07:13,  1.45it/s]\u001b[A\n",
      " 56%|██████████████████████▏                 | 787/1415 [08:16<07:13,  1.45it/s]\u001b[A\n",
      " 56%|██████████████████████▎                 | 788/1415 [08:17<07:13,  1.45it/s]\u001b[A\n",
      " 56%|██████████████████████▎                 | 789/1415 [08:18<07:13,  1.44it/s]\u001b[A\n",
      " 56%|██████████████████████▎                 | 790/1415 [08:18<07:12,  1.44it/s]\u001b[A\n",
      " 56%|██████████████████████▎                 | 791/1415 [08:19<07:11,  1.44it/s]\u001b[A\n",
      " 56%|██████████████████████▍                 | 792/1415 [08:20<07:12,  1.44it/s]\u001b[A\n",
      " 56%|██████████████████████▍                 | 793/1415 [08:20<07:14,  1.43it/s]\u001b[A\n",
      " 56%|██████████████████████▍                 | 794/1415 [08:21<07:06,  1.46it/s]\u001b[A\n",
      " 56%|██████████████████████▍                 | 795/1415 [08:22<07:12,  1.43it/s]\u001b[A\n",
      " 56%|██████████████████████▌                 | 796/1415 [08:23<07:01,  1.47it/s]\u001b[A\n",
      " 56%|██████████████████████▌                 | 797/1415 [08:23<07:01,  1.46it/s]\u001b[A\n",
      " 56%|██████████████████████▌                 | 798/1415 [08:24<06:53,  1.49it/s]\u001b[A\n",
      " 56%|██████████████████████▌                 | 799/1415 [08:25<07:01,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▌                 | 800/1415 [08:25<07:00,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▋                 | 801/1415 [08:26<07:00,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▋                 | 802/1415 [08:27<07:07,  1.44it/s]\u001b[A\n",
      " 57%|██████████████████████▋                 | 803/1415 [08:27<06:59,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▋                 | 804/1415 [08:28<06:57,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▊                 | 805/1415 [08:29<06:56,  1.47it/s]\u001b[A\n",
      " 57%|██████████████████████▊                 | 806/1415 [08:29<07:02,  1.44it/s]\u001b[A\n",
      " 57%|██████████████████████▊                 | 807/1415 [08:30<06:58,  1.45it/s]\u001b[A\n",
      " 57%|██████████████████████▊                 | 808/1415 [08:31<07:02,  1.44it/s]\u001b[A\n",
      " 57%|██████████████████████▊                 | 809/1415 [08:31<07:01,  1.44it/s]\u001b[A\n",
      " 57%|██████████████████████▉                 | 810/1415 [08:32<06:55,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▉                 | 811/1415 [08:33<06:53,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▉                 | 812/1415 [08:34<06:51,  1.46it/s]\u001b[A\n",
      " 57%|██████████████████████▉                 | 813/1415 [08:34<07:03,  1.42it/s]\u001b[A\n",
      " 58%|███████████████████████                 | 814/1415 [08:35<06:57,  1.44it/s]\u001b[A\n",
      " 58%|███████████████████████                 | 815/1415 [08:36<06:58,  1.43it/s]\u001b[A\n",
      " 58%|███████████████████████                 | 816/1415 [08:36<06:52,  1.45it/s]\u001b[A\n",
      " 58%|███████████████████████                 | 817/1415 [08:37<06:49,  1.46it/s]\u001b[A\n",
      " 58%|███████████████████████                 | 818/1415 [08:38<06:47,  1.46it/s]\u001b[A\n",
      " 58%|███████████████████████▏                | 819/1415 [08:38<06:48,  1.46it/s]\u001b[A\n",
      " 58%|███████████████████████▏                | 820/1415 [08:39<06:56,  1.43it/s]\u001b[A\n",
      " 58%|███████████████████████▏                | 821/1415 [08:40<06:53,  1.43it/s]\u001b[A\n",
      " 58%|███████████████████████▏                | 822/1415 [08:40<06:53,  1.43it/s]\u001b[A\n",
      " 58%|███████████████████████▎                | 823/1415 [08:41<06:42,  1.47it/s]\u001b[A\n",
      " 58%|███████████████████████▎                | 824/1415 [08:42<06:41,  1.47it/s]\u001b[A\n",
      " 58%|███████████████████████▎                | 825/1415 [08:42<06:37,  1.48it/s]\u001b[A\n",
      " 58%|███████████████████████▎                | 826/1415 [08:43<06:44,  1.46it/s]\u001b[A\n",
      " 58%|███████████████████████▍                | 827/1415 [08:44<06:42,  1.46it/s]\u001b[A\n",
      " 59%|███████████████████████▍                | 828/1415 [08:45<06:42,  1.46it/s]\u001b[A\n",
      " 59%|███████████████████████▍                | 829/1415 [08:45<06:42,  1.46it/s]\u001b[A\n",
      " 59%|███████████████████████▍                | 830/1415 [08:46<06:33,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▍                | 831/1415 [08:47<06:40,  1.46it/s]\u001b[A\n",
      " 59%|███████████████████████▌                | 832/1415 [08:47<06:32,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▌                | 833/1415 [08:48<06:31,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▌                | 834/1415 [08:49<06:25,  1.51it/s]\u001b[A\n",
      " 59%|███████████████████████▌                | 835/1415 [08:49<06:29,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▋                | 836/1415 [08:50<06:32,  1.47it/s]\u001b[A\n",
      " 59%|███████████████████████▋                | 837/1415 [08:51<06:32,  1.47it/s]\u001b[A\n",
      " 59%|███████████████████████▋                | 838/1415 [08:51<06:34,  1.46it/s]\u001b[A\n",
      " 59%|███████████████████████▋                | 839/1415 [08:52<06:26,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▋                | 840/1415 [08:53<06:26,  1.49it/s]\u001b[A\n",
      " 59%|███████████████████████▊                | 841/1415 [08:53<06:20,  1.51it/s]\u001b[A\n",
      " 60%|███████████████████████▊                | 842/1415 [08:54<06:22,  1.50it/s]\u001b[A\n",
      " 60%|███████████████████████▊                | 843/1415 [08:55<06:19,  1.51it/s]\u001b[A\n",
      " 60%|███████████████████████▊                | 844/1415 [08:55<06:19,  1.50it/s]\u001b[A\n",
      " 60%|███████████████████████▉                | 845/1415 [08:56<06:20,  1.50it/s]\u001b[A\n",
      " 60%|███████████████████████▉                | 846/1415 [08:57<06:24,  1.48it/s]\u001b[A\n",
      " 60%|███████████████████████▉                | 847/1415 [08:57<06:31,  1.45it/s]\u001b[A\n",
      " 60%|███████████████████████▉                | 848/1415 [08:58<06:23,  1.48it/s]\u001b[A\n",
      " 60%|████████████████████████                | 849/1415 [08:59<06:29,  1.45it/s]\u001b[A\n",
      " 60%|████████████████████████                | 850/1415 [08:59<06:30,  1.45it/s]\u001b[A\n",
      " 60%|████████████████████████                | 851/1415 [09:00<06:28,  1.45it/s]\u001b[A\n",
      " 60%|████████████████████████                | 852/1415 [09:01<06:28,  1.45it/s]\u001b[A\n",
      " 60%|████████████████████████                | 853/1415 [09:01<06:31,  1.44it/s]\u001b[A\n",
      " 60%|████████████████████████▏               | 854/1415 [09:02<06:35,  1.42it/s]\u001b[A\n",
      " 60%|████████████████████████▏               | 855/1415 [09:03<06:24,  1.46it/s]\u001b[A\n",
      " 60%|████████████████████████▏               | 856/1415 [09:04<06:29,  1.44it/s]\u001b[A\n",
      " 61%|████████████████████████▏               | 857/1415 [09:04<06:21,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▎               | 858/1415 [09:05<06:20,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▎               | 859/1415 [09:06<06:21,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▎               | 860/1415 [09:06<06:20,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▎               | 861/1415 [09:07<06:16,  1.47it/s]\u001b[A\n",
      " 61%|████████████████████████▎               | 862/1415 [09:08<06:27,  1.43it/s]\u001b[A\n",
      " 61%|████████████████████████▍               | 863/1415 [09:08<06:28,  1.42it/s]\u001b[A\n",
      " 61%|████████████████████████▍               | 864/1415 [09:09<06:18,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▍               | 865/1415 [09:10<06:25,  1.43it/s]\u001b[A\n",
      " 61%|████████████████████████▍               | 866/1415 [09:10<06:23,  1.43it/s]\u001b[A\n",
      " 61%|████████████████████████▌               | 867/1415 [09:11<06:21,  1.43it/s]\u001b[A\n",
      " 61%|████████████████████████▌               | 868/1415 [09:12<06:13,  1.46it/s]\u001b[A\n",
      " 61%|████████████████████████▌               | 869/1415 [09:13<06:19,  1.44it/s]\u001b[A\n",
      " 61%|████████████████████████▌               | 870/1415 [09:13<06:17,  1.44it/s]\u001b[A\n",
      " 62%|████████████████████████▌               | 871/1415 [09:14<06:24,  1.41it/s]\u001b[A\n",
      " 62%|████████████████████████▋               | 872/1415 [09:15<06:26,  1.41it/s]\u001b[A\n",
      " 62%|████████████████████████▋               | 873/1415 [09:15<06:22,  1.42it/s]\u001b[A\n",
      " 62%|████████████████████████▋               | 874/1415 [09:16<06:25,  1.40it/s]\u001b[A\n",
      " 62%|████████████████████████▋               | 875/1415 [09:17<06:16,  1.44it/s]\u001b[A\n",
      " 62%|████████████████████████▊               | 876/1415 [09:18<06:20,  1.42it/s]\u001b[A\n",
      " 62%|████████████████████████▊               | 877/1415 [09:18<06:11,  1.45it/s]\u001b[A\n",
      " 62%|████████████████████████▊               | 878/1415 [09:19<06:19,  1.42it/s]\u001b[A\n",
      " 62%|████████████████████████▊               | 879/1415 [09:20<06:11,  1.44it/s]\u001b[A\n",
      " 62%|████████████████████████▉               | 880/1415 [09:20<06:13,  1.43it/s]\u001b[A\n",
      " 62%|████████████████████████▉               | 881/1415 [09:21<06:21,  1.40it/s]\u001b[A\n",
      " 62%|████████████████████████▉               | 882/1415 [09:22<06:11,  1.44it/s]\u001b[A\n",
      " 62%|████████████████████████▉               | 883/1415 [09:22<06:10,  1.44it/s]\u001b[A\n",
      " 62%|████████████████████████▉               | 884/1415 [09:23<06:04,  1.46it/s]\u001b[A\n",
      " 63%|█████████████████████████               | 885/1415 [09:24<06:12,  1.42it/s]\u001b[A\n",
      " 63%|█████████████████████████               | 886/1415 [09:25<06:12,  1.42it/s]\u001b[A\n",
      " 63%|█████████████████████████               | 887/1415 [09:25<06:16,  1.40it/s]\u001b[A\n",
      " 63%|█████████████████████████               | 888/1415 [09:26<06:13,  1.41it/s]\u001b[A\n",
      " 63%|█████████████████████████▏              | 889/1415 [09:27<06:16,  1.40it/s]\u001b[A\n",
      " 63%|█████████████████████████▏              | 890/1415 [09:27<06:24,  1.36it/s]\u001b[A\n",
      " 63%|█████████████████████████▏              | 891/1415 [09:28<06:20,  1.38it/s]\u001b[A\n",
      " 63%|█████████████████████████▏              | 892/1415 [09:29<06:15,  1.39it/s]\u001b[A\n",
      " 63%|█████████████████████████▏              | 893/1415 [09:30<06:10,  1.41it/s]\u001b[A\n",
      " 63%|█████████████████████████▎              | 894/1415 [09:30<06:15,  1.39it/s]\u001b[A\n",
      " 63%|█████████████████████████▎              | 895/1415 [09:31<06:02,  1.43it/s]\u001b[A\n",
      " 63%|█████████████████████████▎              | 896/1415 [09:32<06:00,  1.44it/s]\u001b[A\n",
      " 63%|█████████████████████████▎              | 897/1415 [09:32<05:58,  1.45it/s]\u001b[A\n",
      " 63%|█████████████████████████▍              | 898/1415 [09:33<05:56,  1.45it/s]\u001b[A\n",
      " 64%|█████████████████████████▍              | 899/1415 [09:34<05:55,  1.45it/s]\u001b[A\n",
      " 64%|█████████████████████████▍              | 900/1415 [09:34<05:50,  1.47it/s]\u001b[A\n",
      " 64%|█████████████████████████▍              | 901/1415 [09:35<05:50,  1.47it/s]\u001b[A\n",
      " 64%|█████████████████████████▍              | 902/1415 [09:36<05:44,  1.49it/s]\u001b[A\n",
      " 64%|█████████████████████████▌              | 903/1415 [09:36<05:50,  1.46it/s]\u001b[A\n",
      " 64%|█████████████████████████▌              | 904/1415 [09:37<05:49,  1.46it/s]\u001b[A\n",
      " 64%|█████████████████████████▌              | 905/1415 [09:38<05:49,  1.46it/s]\u001b[A\n",
      " 64%|█████████████████████████▌              | 906/1415 [09:38<05:44,  1.48it/s]\u001b[A\n",
      " 64%|█████████████████████████▋              | 907/1415 [09:39<05:45,  1.47it/s]\u001b[A\n",
      " 64%|█████████████████████████▋              | 908/1415 [09:40<05:46,  1.46it/s]\u001b[A\n",
      " 64%|█████████████████████████▋              | 909/1415 [09:40<05:43,  1.47it/s]\u001b[A\n",
      " 64%|█████████████████████████▋              | 910/1415 [09:41<05:45,  1.46it/s]\u001b[A\n",
      " 64%|█████████████████████████▊              | 911/1415 [09:42<05:42,  1.47it/s]\u001b[A\n",
      " 64%|█████████████████████████▊              | 912/1415 [09:43<05:41,  1.47it/s]\u001b[A\n",
      " 65%|█████████████████████████▊              | 913/1415 [09:43<05:39,  1.48it/s]\u001b[A\n",
      " 65%|█████████████████████████▊              | 914/1415 [09:44<05:43,  1.46it/s]\u001b[A\n",
      " 65%|█████████████████████████▊              | 915/1415 [09:45<05:40,  1.47it/s]\u001b[A\n",
      " 65%|█████████████████████████▉              | 916/1415 [09:45<05:41,  1.46it/s]\u001b[A\n",
      " 65%|█████████████████████████▉              | 917/1415 [09:46<05:47,  1.43it/s]\u001b[A\n",
      " 65%|█████████████████████████▉              | 918/1415 [09:47<05:40,  1.46it/s]\u001b[A\n",
      " 65%|█████████████████████████▉              | 919/1415 [09:47<05:40,  1.46it/s]\u001b[A\n",
      " 65%|██████████████████████████              | 920/1415 [09:48<05:31,  1.49it/s]\u001b[A\n",
      " 65%|██████████████████████████              | 921/1415 [09:49<05:28,  1.50it/s]\u001b[A\n",
      " 65%|██████████████████████████              | 922/1415 [09:49<05:27,  1.51it/s]\u001b[A\n",
      " 65%|██████████████████████████              | 923/1415 [09:50<05:26,  1.51it/s]\u001b[A\n",
      " 65%|██████████████████████████              | 924/1415 [09:51<05:25,  1.51it/s]\u001b[A\n",
      " 65%|██████████████████████████▏             | 925/1415 [09:51<05:32,  1.48it/s]\u001b[A\n",
      " 65%|██████████████████████████▏             | 926/1415 [09:52<05:30,  1.48it/s]\u001b[A\n",
      " 66%|██████████████████████████▏             | 927/1415 [09:53<05:30,  1.48it/s]\u001b[A\n",
      " 66%|██████████████████████████▏             | 928/1415 [09:53<05:28,  1.48it/s]\u001b[A\n",
      " 66%|██████████████████████████▎             | 929/1415 [09:54<05:22,  1.51it/s]\u001b[A\n",
      " 66%|██████████████████████████▎             | 930/1415 [09:55<05:27,  1.48it/s]\u001b[A\n",
      " 66%|██████████████████████████▎             | 931/1415 [09:55<05:18,  1.52it/s]\u001b[A\n",
      " 66%|██████████████████████████▎             | 932/1415 [09:56<05:22,  1.50it/s]\u001b[A\n",
      " 66%|██████████████████████████▎             | 933/1415 [09:57<05:13,  1.54it/s]\u001b[A\n",
      " 66%|██████████████████████████▍             | 934/1415 [09:57<05:12,  1.54it/s]\u001b[A\n",
      " 66%|██████████████████████████▍             | 935/1415 [09:58<05:10,  1.54it/s]\u001b[A\n",
      " 66%|██████████████████████████▍             | 936/1415 [09:59<05:10,  1.54it/s]\u001b[A\n",
      " 66%|██████████████████████████▍             | 937/1415 [09:59<05:12,  1.53it/s]\u001b[A\n",
      " 66%|██████████████████████████▌             | 938/1415 [10:00<05:05,  1.56it/s]\u001b[A\n",
      " 66%|██████████████████████████▌             | 939/1415 [10:00<05:07,  1.55it/s]\u001b[A\n",
      " 66%|██████████████████████████▌             | 940/1415 [10:01<05:03,  1.57it/s]\u001b[A\n",
      " 67%|██████████████████████████▌             | 941/1415 [10:02<05:02,  1.57it/s]\u001b[A\n",
      " 67%|██████████████████████████▋             | 942/1415 [10:02<04:59,  1.58it/s]\u001b[A\n",
      " 67%|██████████████████████████▋             | 943/1415 [10:03<05:01,  1.57it/s]\u001b[A\n",
      " 67%|██████████████████████████▋             | 944/1415 [10:04<04:59,  1.57it/s]\u001b[A\n",
      " 67%|██████████████████████████▋             | 945/1415 [10:04<04:55,  1.59it/s]\u001b[A\n",
      " 67%|██████████████████████████▋             | 946/1415 [10:05<04:56,  1.58it/s]\u001b[A\n",
      " 67%|██████████████████████████▊             | 947/1415 [10:05<04:53,  1.59it/s]\u001b[A\n",
      " 67%|██████████████████████████▊             | 948/1415 [10:06<04:55,  1.58it/s]\u001b[A\n",
      " 67%|██████████████████████████▊             | 949/1415 [10:07<04:49,  1.61it/s]\u001b[A\n",
      " 67%|██████████████████████████▊             | 950/1415 [10:07<04:51,  1.59it/s]\u001b[A\n",
      " 67%|██████████████████████████▉             | 951/1415 [10:08<04:49,  1.60it/s]\u001b[A\n",
      " 67%|██████████████████████████▉             | 952/1415 [10:09<04:50,  1.59it/s]\u001b[A\n",
      " 67%|██████████████████████████▉             | 953/1415 [10:09<04:53,  1.57it/s]\u001b[A\n",
      " 67%|██████████████████████████▉             | 954/1415 [10:10<04:55,  1.56it/s]\u001b[A\n",
      " 67%|██████████████████████████▉             | 955/1415 [10:11<04:58,  1.54it/s]\u001b[A\n",
      " 68%|███████████████████████████             | 956/1415 [10:11<04:54,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████             | 957/1415 [10:12<04:53,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████             | 958/1415 [10:12<04:48,  1.58it/s]\u001b[A\n",
      " 68%|███████████████████████████             | 959/1415 [10:13<04:49,  1.57it/s]\u001b[A\n",
      " 68%|███████████████████████████▏            | 960/1415 [10:14<04:45,  1.59it/s]\u001b[A\n",
      " 68%|███████████████████████████▏            | 961/1415 [10:14<04:50,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████▏            | 962/1415 [10:15<04:51,  1.55it/s]\u001b[A\n",
      " 68%|███████████████████████████▏            | 963/1415 [10:16<04:47,  1.57it/s]\u001b[A\n",
      " 68%|███████████████████████████▎            | 964/1415 [10:16<04:49,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████▎            | 965/1415 [10:17<04:49,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████▎            | 966/1415 [10:18<04:47,  1.56it/s]\u001b[A\n",
      " 68%|███████████████████████████▎            | 967/1415 [10:18<04:43,  1.58it/s]\u001b[A\n",
      " 68%|███████████████████████████▎            | 968/1415 [10:19<04:48,  1.55it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 68%|███████████████████████████▍            | 969/1415 [10:20<04:47,  1.55it/s]\u001b[A\n",
      " 69%|███████████████████████████▍            | 970/1415 [10:20<04:47,  1.55it/s]\u001b[A\n",
      " 69%|███████████████████████████▍            | 971/1415 [10:21<04:45,  1.56it/s]\u001b[A\n",
      " 69%|███████████████████████████▍            | 972/1415 [10:21<04:40,  1.58it/s]\u001b[A\n",
      " 69%|███████████████████████████▌            | 973/1415 [10:22<04:42,  1.57it/s]\u001b[A\n",
      " 69%|███████████████████████████▌            | 974/1415 [10:23<04:37,  1.59it/s]\u001b[A\n",
      " 69%|███████████████████████████▌            | 975/1415 [10:23<04:39,  1.58it/s]\u001b[A\n",
      " 69%|███████████████████████████▌            | 976/1415 [10:24<04:35,  1.59it/s]\u001b[A\n",
      " 69%|███████████████████████████▌            | 977/1415 [10:25<04:42,  1.55it/s]\u001b[A\n",
      " 69%|███████████████████████████▋            | 978/1415 [10:25<04:41,  1.55it/s]\u001b[A\n",
      " 69%|███████████████████████████▋            | 979/1415 [10:26<04:40,  1.56it/s]\u001b[A\n",
      " 69%|███████████████████████████▋            | 980/1415 [10:27<04:40,  1.55it/s]\u001b[A\n",
      " 69%|███████████████████████████▋            | 981/1415 [10:27<04:38,  1.56it/s]\u001b[A\n",
      " 69%|███████████████████████████▊            | 982/1415 [10:28<04:37,  1.56it/s]\u001b[A\n",
      " 69%|███████████████████████████▊            | 983/1415 [10:28<04:37,  1.55it/s]\u001b[A\n",
      " 70%|███████████████████████████▊            | 984/1415 [10:29<04:37,  1.55it/s]\u001b[A\n",
      " 70%|███████████████████████████▊            | 985/1415 [10:30<04:32,  1.58it/s]\u001b[A\n",
      " 70%|███████████████████████████▊            | 986/1415 [10:30<04:37,  1.55it/s]\u001b[A\n",
      " 70%|███████████████████████████▉            | 987/1415 [10:31<04:32,  1.57it/s]\u001b[A\n",
      " 70%|███████████████████████████▉            | 988/1415 [10:32<04:31,  1.57it/s]\u001b[A\n",
      " 70%|███████████████████████████▉            | 989/1415 [10:32<04:30,  1.57it/s]\u001b[A\n",
      " 70%|███████████████████████████▉            | 990/1415 [10:33<04:26,  1.60it/s]\u001b[A\n",
      " 70%|████████████████████████████            | 991/1415 [10:34<04:27,  1.59it/s]\u001b[A\n",
      " 70%|████████████████████████████            | 992/1415 [10:34<04:23,  1.60it/s]\u001b[A\n",
      " 70%|████████████████████████████            | 993/1415 [10:35<04:31,  1.55it/s]\u001b[A\n",
      " 70%|████████████████████████████            | 994/1415 [10:35<04:27,  1.57it/s]\u001b[A\n",
      " 70%|████████████████████████████▏           | 995/1415 [10:36<04:27,  1.57it/s]\u001b[A\n",
      " 70%|████████████████████████████▏           | 996/1415 [10:37<04:22,  1.59it/s]\u001b[A\n",
      " 70%|████████████████████████████▏           | 997/1415 [10:37<04:28,  1.56it/s]\u001b[A\n",
      " 71%|████████████████████████████▏           | 998/1415 [10:38<04:32,  1.53it/s]\u001b[A\n",
      " 71%|████████████████████████████▏           | 999/1415 [10:39<04:26,  1.56it/s]\u001b[A\n",
      " 71%|███████████████████████████▌           | 1000/1415 [10:39<04:28,  1.54it/s]\u001b[A\n",
      " 71%|███████████████████████████▌           | 1001/1415 [10:40<04:24,  1.57it/s]\u001b[A\n",
      " 71%|███████████████████████████▌           | 1002/1415 [10:41<04:29,  1.53it/s]\u001b[A\n",
      " 71%|███████████████████████████▋           | 1003/1415 [10:41<04:24,  1.56it/s]\u001b[A\n",
      " 71%|███████████████████████████▋           | 1004/1415 [10:42<04:29,  1.53it/s]\u001b[A\n",
      " 71%|███████████████████████████▋           | 1005/1415 [10:43<04:25,  1.54it/s]\u001b[A\n",
      " 71%|███████████████████████████▋           | 1006/1415 [10:43<04:26,  1.53it/s]\u001b[A\n",
      " 71%|███████████████████████████▊           | 1007/1415 [10:44<04:25,  1.54it/s]\u001b[A\n",
      " 71%|███████████████████████████▊           | 1008/1415 [10:45<04:25,  1.53it/s]\u001b[A\n",
      " 71%|███████████████████████████▊           | 1009/1415 [10:45<04:25,  1.53it/s]\u001b[A\n",
      " 71%|███████████████████████████▊           | 1010/1415 [10:46<04:20,  1.56it/s]\u001b[A\n",
      " 71%|███████████████████████████▊           | 1011/1415 [10:46<04:19,  1.56it/s]\u001b[A\n",
      " 72%|███████████████████████████▉           | 1012/1415 [10:47<04:15,  1.58it/s]\u001b[A\n",
      " 72%|███████████████████████████▉           | 1013/1415 [10:48<04:15,  1.57it/s]\u001b[A\n",
      " 72%|███████████████████████████▉           | 1014/1415 [10:48<04:11,  1.59it/s]\u001b[A\n",
      " 72%|███████████████████████████▉           | 1015/1415 [10:49<04:13,  1.58it/s]\u001b[A\n",
      " 72%|████████████████████████████           | 1016/1415 [10:50<04:20,  1.53it/s]\u001b[A\n",
      " 72%|████████████████████████████           | 1017/1415 [10:50<04:16,  1.55it/s]\u001b[A\n",
      " 72%|████████████████████████████           | 1018/1415 [10:51<04:15,  1.55it/s]\u001b[A\n",
      " 72%|████████████████████████████           | 1019/1415 [10:52<04:13,  1.56it/s]\u001b[A\n",
      " 72%|████████████████████████████           | 1020/1415 [10:52<04:16,  1.54it/s]\u001b[A\n",
      " 72%|████████████████████████████▏          | 1021/1415 [10:53<04:11,  1.57it/s]\u001b[A\n",
      " 72%|████████████████████████████▏          | 1022/1415 [10:54<04:16,  1.53it/s]\u001b[A\n",
      " 72%|████████████████████████████▏          | 1023/1415 [10:54<04:15,  1.53it/s]\u001b[A\n",
      " 72%|████████████████████████████▏          | 1024/1415 [10:55<04:13,  1.54it/s]\u001b[A\n",
      " 72%|████████████████████████████▎          | 1025/1415 [10:55<04:11,  1.55it/s]\u001b[A\n",
      " 73%|████████████████████████████▎          | 1026/1415 [10:56<04:06,  1.58it/s]\u001b[A\n",
      " 73%|████████████████████████████▎          | 1027/1415 [10:57<04:07,  1.57it/s]\u001b[A\n",
      " 73%|████████████████████████████▎          | 1028/1415 [10:57<04:05,  1.58it/s]\u001b[A\n",
      " 73%|████████████████████████████▎          | 1029/1415 [10:58<04:05,  1.57it/s]\u001b[A\n",
      " 73%|████████████████████████████▍          | 1030/1415 [10:59<04:02,  1.59it/s]\u001b[A\n",
      " 73%|████████████████████████████▍          | 1031/1415 [10:59<04:07,  1.55it/s]\u001b[A\n",
      " 73%|████████████████████████████▍          | 1032/1415 [11:00<04:08,  1.54it/s]\u001b[A\n",
      " 73%|████████████████████████████▍          | 1033/1415 [11:01<04:07,  1.54it/s]\u001b[A\n",
      " 73%|████████████████████████████▍          | 1034/1415 [11:01<04:06,  1.55it/s]\u001b[A\n",
      " 73%|████████████████████████████▌          | 1035/1415 [11:02<04:00,  1.58it/s]\u001b[A\n",
      " 73%|████████████████████████████▌          | 1036/1415 [11:02<04:01,  1.57it/s]\u001b[A\n",
      " 73%|████████████████████████████▌          | 1037/1415 [11:03<03:57,  1.59it/s]\u001b[A\n",
      " 73%|████████████████████████████▌          | 1038/1415 [11:04<04:04,  1.54it/s]\u001b[A\n",
      " 73%|████████████████████████████▋          | 1039/1415 [11:04<03:59,  1.57it/s]\u001b[A\n",
      " 73%|████████████████████████████▋          | 1040/1415 [11:05<03:59,  1.57it/s]\u001b[A\n",
      " 74%|████████████████████████████▋          | 1041/1415 [11:06<03:55,  1.59it/s]\u001b[A\n",
      " 74%|████████████████████████████▋          | 1042/1415 [11:06<03:59,  1.56it/s]\u001b[A\n",
      " 74%|████████████████████████████▋          | 1043/1415 [11:07<03:58,  1.56it/s]\u001b[A\n",
      " 74%|████████████████████████████▊          | 1044/1415 [11:08<03:54,  1.59it/s]\u001b[A\n",
      " 74%|████████████████████████████▊          | 1045/1415 [11:08<03:55,  1.57it/s]\u001b[A\n",
      " 74%|████████████████████████████▊          | 1046/1415 [11:09<03:51,  1.59it/s]\u001b[A\n",
      " 74%|████████████████████████████▊          | 1047/1415 [11:09<03:52,  1.58it/s]\u001b[A\n",
      " 74%|████████████████████████████▉          | 1048/1415 [11:10<03:54,  1.57it/s]\u001b[A\n",
      " 74%|████████████████████████████▉          | 1049/1415 [11:11<03:58,  1.54it/s]\u001b[A\n",
      " 74%|████████████████████████████▉          | 1050/1415 [11:11<03:55,  1.55it/s]\u001b[A\n",
      " 74%|████████████████████████████▉          | 1051/1415 [11:12<03:54,  1.55it/s]\u001b[A\n",
      " 74%|████████████████████████████▉          | 1052/1415 [11:13<03:59,  1.52it/s]\u001b[A\n",
      " 74%|█████████████████████████████          | 1053/1415 [11:13<03:56,  1.53it/s]\u001b[A\n",
      " 74%|█████████████████████████████          | 1054/1415 [11:14<03:56,  1.52it/s]\u001b[A\n",
      " 75%|█████████████████████████████          | 1055/1415 [11:15<03:50,  1.56it/s]\u001b[A\n",
      " 75%|█████████████████████████████          | 1056/1415 [11:15<03:50,  1.56it/s]\u001b[A\n",
      " 75%|█████████████████████████████▏         | 1057/1415 [11:16<03:47,  1.58it/s]\u001b[A\n",
      " 75%|█████████████████████████████▏         | 1058/1415 [11:17<03:51,  1.54it/s]\u001b[A\n",
      " 75%|█████████████████████████████▏         | 1059/1415 [11:17<03:47,  1.56it/s]\u001b[A\n",
      " 75%|█████████████████████████████▏         | 1060/1415 [11:18<03:46,  1.57it/s]\u001b[A\n",
      " 75%|█████████████████████████████▏         | 1061/1415 [11:19<03:49,  1.54it/s]\u001b[A\n",
      " 75%|█████████████████████████████▎         | 1062/1415 [11:19<03:46,  1.56it/s]\u001b[A\n",
      " 75%|█████████████████████████████▎         | 1063/1415 [11:20<03:51,  1.52it/s]\u001b[A\n",
      " 75%|█████████████████████████████▎         | 1064/1415 [11:20<03:51,  1.52it/s]\u001b[A\n",
      " 75%|█████████████████████████████▎         | 1065/1415 [11:21<03:52,  1.51it/s]\u001b[A\n",
      " 75%|█████████████████████████████▍         | 1066/1415 [11:22<03:50,  1.51it/s]\u001b[A\n",
      " 75%|█████████████████████████████▍         | 1067/1415 [11:22<03:51,  1.50it/s]\u001b[A\n",
      " 75%|█████████████████████████████▍         | 1068/1415 [11:23<03:48,  1.52it/s]\u001b[A\n",
      " 76%|█████████████████████████████▍         | 1069/1415 [11:24<03:50,  1.50it/s]\u001b[A\n",
      " 76%|█████████████████████████████▍         | 1070/1415 [11:25<03:53,  1.48it/s]\u001b[A\n",
      " 76%|█████████████████████████████▌         | 1071/1415 [11:25<03:50,  1.49it/s]\u001b[A\n",
      " 76%|█████████████████████████████▌         | 1072/1415 [11:26<03:51,  1.48it/s]\u001b[A\n",
      " 76%|█████████████████████████████▌         | 1073/1415 [11:27<03:48,  1.50it/s]\u001b[A\n",
      " 76%|█████████████████████████████▌         | 1074/1415 [11:27<03:48,  1.49it/s]\u001b[A\n",
      " 76%|█████████████████████████████▋         | 1075/1415 [11:28<03:41,  1.53it/s]\u001b[A\n",
      " 76%|█████████████████████████████▋         | 1076/1415 [11:28<03:39,  1.54it/s]\u001b[A\n",
      " 76%|█████████████████████████████▋         | 1077/1415 [11:29<03:35,  1.57it/s]\u001b[A\n",
      " 76%|█████████████████████████████▋         | 1078/1415 [11:30<03:40,  1.53it/s]\u001b[A\n",
      " 76%|█████████████████████████████▋         | 1079/1415 [11:30<03:43,  1.50it/s]\u001b[A\n",
      " 76%|█████████████████████████████▊         | 1080/1415 [11:31<03:41,  1.51it/s]\u001b[A\n",
      " 76%|█████████████████████████████▊         | 1081/1415 [11:32<03:39,  1.52it/s]\u001b[A\n",
      " 76%|█████████████████████████████▊         | 1082/1415 [11:32<03:34,  1.55it/s]\u001b[A\n",
      " 77%|█████████████████████████████▊         | 1083/1415 [11:33<03:38,  1.52it/s]\u001b[A\n",
      " 77%|█████████████████████████████▉         | 1084/1415 [11:34<03:37,  1.52it/s]\u001b[A\n",
      " 77%|█████████████████████████████▉         | 1085/1415 [11:34<03:41,  1.49it/s]\u001b[A\n",
      " 77%|█████████████████████████████▉         | 1086/1415 [11:35<03:39,  1.50it/s]\u001b[A\n",
      " 77%|█████████████████████████████▉         | 1087/1415 [11:36<03:35,  1.53it/s]\u001b[A\n",
      " 77%|█████████████████████████████▉         | 1088/1415 [11:36<03:35,  1.52it/s]\u001b[A\n",
      " 77%|██████████████████████████████         | 1089/1415 [11:37<03:29,  1.55it/s]\u001b[A\n",
      " 77%|██████████████████████████████         | 1090/1415 [11:38<03:31,  1.54it/s]\u001b[A\n",
      " 77%|██████████████████████████████         | 1091/1415 [11:38<03:26,  1.57it/s]\u001b[A\n",
      " 77%|██████████████████████████████         | 1092/1415 [11:39<03:28,  1.55it/s]\u001b[A\n",
      " 77%|██████████████████████████████▏        | 1093/1415 [11:40<03:25,  1.57it/s]\u001b[A\n",
      " 77%|██████████████████████████████▏        | 1094/1415 [11:40<03:25,  1.56it/s]\u001b[A\n",
      " 77%|██████████████████████████████▏        | 1095/1415 [11:41<03:22,  1.58it/s]\u001b[A\n",
      " 77%|██████████████████████████████▏        | 1096/1415 [11:41<03:25,  1.55it/s]\u001b[A\n",
      " 78%|██████████████████████████████▏        | 1097/1415 [11:42<03:25,  1.55it/s]\u001b[A\n",
      " 78%|██████████████████████████████▎        | 1098/1415 [11:43<03:21,  1.57it/s]\u001b[A\n",
      " 78%|██████████████████████████████▎        | 1099/1415 [11:43<03:22,  1.56it/s]\u001b[A\n",
      " 78%|██████████████████████████████▎        | 1100/1415 [11:44<03:19,  1.58it/s]\u001b[A\n",
      " 78%|██████████████████████████████▎        | 1101/1415 [11:45<03:22,  1.55it/s]\u001b[A\n",
      " 78%|██████████████████████████████▎        | 1102/1415 [11:45<03:19,  1.57it/s]\u001b[A\n",
      " 78%|██████████████████████████████▍        | 1103/1415 [11:46<03:18,  1.57it/s]\u001b[A\n",
      " 78%|██████████████████████████████▍        | 1104/1415 [11:47<03:19,  1.56it/s]\u001b[A\n",
      " 78%|██████████████████████████████▍        | 1105/1415 [11:47<03:20,  1.54it/s]\u001b[A\n",
      " 78%|██████████████████████████████▍        | 1106/1415 [11:48<03:20,  1.54it/s]\u001b[A\n",
      " 78%|██████████████████████████████▌        | 1107/1415 [11:49<03:23,  1.52it/s]\u001b[A\n",
      " 78%|██████████████████████████████▌        | 1108/1415 [11:49<03:26,  1.48it/s]\u001b[A\n",
      " 78%|██████████████████████████████▌        | 1109/1415 [11:50<03:25,  1.49it/s]\u001b[A\n",
      " 78%|██████████████████████████████▌        | 1110/1415 [11:51<03:26,  1.48it/s]\u001b[A\n",
      " 79%|██████████████████████████████▌        | 1111/1415 [11:51<03:20,  1.51it/s]\u001b[A\n",
      " 79%|██████████████████████████████▋        | 1112/1415 [11:52<03:18,  1.53it/s]\u001b[A\n",
      " 79%|██████████████████████████████▋        | 1113/1415 [11:52<03:14,  1.56it/s]\u001b[A\n",
      " 79%|██████████████████████████████▋        | 1114/1415 [11:53<03:13,  1.56it/s]\u001b[A\n",
      " 79%|██████████████████████████████▋        | 1115/1415 [11:54<03:11,  1.56it/s]\u001b[A\n",
      " 79%|██████████████████████████████▊        | 1116/1415 [11:54<03:11,  1.57it/s]\u001b[A\n",
      " 79%|██████████████████████████████▊        | 1117/1415 [11:55<03:11,  1.55it/s]\u001b[A\n",
      " 79%|██████████████████████████████▊        | 1118/1415 [11:56<03:11,  1.55it/s]\u001b[A\n",
      " 79%|██████████████████████████████▊        | 1119/1415 [11:56<03:10,  1.55it/s]\u001b[A\n",
      " 79%|██████████████████████████████▊        | 1120/1415 [11:57<03:06,  1.58it/s]\u001b[A\n",
      " 79%|██████████████████████████████▉        | 1121/1415 [11:58<03:05,  1.58it/s]\u001b[A\n",
      " 79%|██████████████████████████████▉        | 1122/1415 [11:58<03:02,  1.60it/s]\u001b[A\n",
      " 79%|██████████████████████████████▉        | 1123/1415 [11:59<03:03,  1.59it/s]\u001b[A\n",
      " 79%|██████████████████████████████▉        | 1124/1415 [11:59<03:03,  1.59it/s]\u001b[A\n",
      " 80%|███████████████████████████████        | 1125/1415 [12:00<03:04,  1.58it/s]\u001b[A\n",
      " 80%|███████████████████████████████        | 1126/1415 [12:01<03:04,  1.57it/s]\u001b[A\n",
      " 80%|███████████████████████████████        | 1127/1415 [12:01<03:03,  1.57it/s]\u001b[A\n",
      " 80%|███████████████████████████████        | 1128/1415 [12:02<03:03,  1.56it/s]\u001b[A\n",
      " 80%|███████████████████████████████        | 1129/1415 [12:03<03:01,  1.57it/s]\u001b[A\n",
      " 80%|███████████████████████████████▏       | 1130/1415 [12:03<03:03,  1.56it/s]\u001b[A\n",
      " 80%|███████████████████████████████▏       | 1131/1415 [12:04<02:59,  1.58it/s]\u001b[A\n",
      " 80%|███████████████████████████████▏       | 1132/1415 [12:05<02:59,  1.58it/s]\u001b[A\n",
      " 80%|███████████████████████████████▏       | 1133/1415 [12:05<03:00,  1.56it/s]\u001b[A\n",
      " 80%|███████████████████████████████▎       | 1134/1415 [12:06<02:56,  1.59it/s]\u001b[A\n",
      " 80%|███████████████████████████████▎       | 1135/1415 [12:06<02:57,  1.58it/s]\u001b[A\n",
      " 80%|███████████████████████████████▎       | 1136/1415 [12:07<02:55,  1.59it/s]\u001b[A\n",
      " 80%|███████████████████████████████▎       | 1137/1415 [12:08<02:54,  1.59it/s]\u001b[A\n",
      " 80%|███████████████████████████████▎       | 1138/1415 [12:08<02:54,  1.59it/s]\u001b[A\n",
      " 80%|███████████████████████████████▍       | 1139/1415 [12:09<02:54,  1.58it/s]\u001b[A\n",
      " 81%|███████████████████████████████▍       | 1140/1415 [12:10<02:52,  1.60it/s]\u001b[A\n",
      " 81%|███████████████████████████████▍       | 1141/1415 [12:10<02:57,  1.55it/s]\u001b[A\n",
      " 81%|███████████████████████████████▍       | 1142/1415 [12:11<02:55,  1.56it/s]\u001b[A\n",
      " 81%|███████████████████████████████▌       | 1143/1415 [12:12<02:55,  1.55it/s]\u001b[A\n",
      " 81%|███████████████████████████████▌       | 1144/1415 [12:12<02:57,  1.53it/s]\u001b[A\n",
      " 81%|███████████████████████████████▌       | 1145/1415 [12:13<02:52,  1.56it/s]\u001b[A\n",
      " 81%|███████████████████████████████▌       | 1146/1415 [12:13<02:51,  1.57it/s]\u001b[A\n",
      " 81%|███████████████████████████████▌       | 1147/1415 [12:14<02:48,  1.59it/s]\u001b[A\n",
      " 81%|███████████████████████████████▋       | 1148/1415 [12:15<02:51,  1.56it/s]\u001b[A\n",
      " 81%|███████████████████████████████▋       | 1149/1415 [12:15<02:51,  1.55it/s]\u001b[A\n",
      " 81%|███████████████████████████████▋       | 1150/1415 [12:16<02:53,  1.53it/s]\u001b[A\n",
      " 81%|███████████████████████████████▋       | 1151/1415 [12:17<02:51,  1.54it/s]\u001b[A\n",
      " 81%|███████████████████████████████▊       | 1152/1415 [12:17<02:47,  1.57it/s]\u001b[A\n",
      " 81%|███████████████████████████████▊       | 1153/1415 [12:18<02:47,  1.57it/s]\u001b[A\n",
      " 82%|███████████████████████████████▊       | 1154/1415 [12:19<02:44,  1.59it/s]\u001b[A\n",
      " 82%|███████████████████████████████▊       | 1155/1415 [12:19<02:43,  1.59it/s]\u001b[A\n",
      " 82%|███████████████████████████████▊       | 1156/1415 [12:20<02:40,  1.61it/s]\u001b[A\n",
      " 82%|███████████████████████████████▉       | 1157/1415 [12:20<02:42,  1.58it/s]\u001b[A\n",
      " 82%|███████████████████████████████▉       | 1158/1415 [12:21<02:41,  1.59it/s]\u001b[A\n",
      " 82%|███████████████████████████████▉       | 1159/1415 [12:22<02:42,  1.57it/s]\u001b[A\n",
      " 82%|███████████████████████████████▉       | 1160/1415 [12:22<02:43,  1.56it/s]\u001b[A\n",
      " 82%|███████████████████████████████▉       | 1161/1415 [12:23<02:40,  1.58it/s]\u001b[A\n",
      " 82%|████████████████████████████████       | 1162/1415 [12:24<02:41,  1.56it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 82%|████████████████████████████████       | 1163/1415 [12:24<02:38,  1.59it/s]\u001b[A\n",
      " 82%|████████████████████████████████       | 1164/1415 [12:25<02:37,  1.59it/s]\u001b[A\n",
      " 82%|████████████████████████████████       | 1165/1415 [12:26<02:34,  1.62it/s]\u001b[A\n",
      " 82%|████████████████████████████████▏      | 1166/1415 [12:26<02:35,  1.60it/s]\u001b[A\n",
      " 82%|████████████████████████████████▏      | 1167/1415 [12:27<02:33,  1.61it/s]\u001b[A\n",
      " 83%|████████████████████████████████▏      | 1168/1415 [12:27<02:35,  1.59it/s]\u001b[A\n",
      " 83%|████████████████████████████████▏      | 1169/1415 [12:28<02:35,  1.59it/s]\u001b[A\n",
      " 83%|████████████████████████████████▏      | 1170/1415 [12:29<02:32,  1.61it/s]\u001b[A\n",
      " 83%|████████████████████████████████▎      | 1171/1415 [12:29<02:33,  1.59it/s]\u001b[A\n",
      " 83%|████████████████████████████████▎      | 1172/1415 [12:30<02:31,  1.60it/s]\u001b[A\n",
      " 83%|████████████████████████████████▎      | 1173/1415 [12:31<02:32,  1.59it/s]\u001b[A\n",
      " 83%|████████████████████████████████▎      | 1174/1415 [12:31<02:30,  1.60it/s]\u001b[A\n",
      " 83%|████████████████████████████████▍      | 1175/1415 [12:32<02:33,  1.56it/s]\u001b[A\n",
      " 83%|████████████████████████████████▍      | 1176/1415 [12:32<02:32,  1.57it/s]\u001b[A\n",
      " 83%|████████████████████████████████▍      | 1177/1415 [12:33<02:31,  1.57it/s]\u001b[A\n",
      " 83%|████████████████████████████████▍      | 1178/1415 [12:34<02:31,  1.57it/s]\u001b[A\n",
      " 83%|████████████████████████████████▍      | 1179/1415 [12:34<02:30,  1.57it/s]\u001b[A\n",
      " 83%|████████████████████████████████▌      | 1180/1415 [12:35<02:29,  1.57it/s]\u001b[A\n",
      " 83%|████████████████████████████████▌      | 1181/1415 [12:36<02:30,  1.56it/s]\u001b[A\n",
      " 84%|████████████████████████████████▌      | 1182/1415 [12:36<02:29,  1.56it/s]\u001b[A\n",
      " 84%|████████████████████████████████▌      | 1183/1415 [12:37<02:25,  1.59it/s]\u001b[A\n",
      " 84%|████████████████████████████████▋      | 1184/1415 [12:38<02:24,  1.60it/s]\u001b[A\n",
      " 84%|████████████████████████████████▋      | 1185/1415 [12:38<02:23,  1.60it/s]\u001b[A\n",
      " 84%|████████████████████████████████▋      | 1186/1415 [12:39<02:22,  1.60it/s]\u001b[A\n",
      " 84%|████████████████████████████████▋      | 1187/1415 [12:39<02:23,  1.59it/s]\u001b[A\n",
      " 84%|████████████████████████████████▋      | 1188/1415 [12:40<02:20,  1.61it/s]\u001b[A\n",
      " 84%|████████████████████████████████▊      | 1189/1415 [12:41<02:24,  1.56it/s]\u001b[A\n",
      " 84%|████████████████████████████████▊      | 1190/1415 [12:41<02:22,  1.58it/s]\u001b[A\n",
      " 84%|████████████████████████████████▊      | 1191/1415 [12:42<02:21,  1.58it/s]\u001b[A\n",
      " 84%|████████████████████████████████▊      | 1192/1415 [12:43<02:19,  1.60it/s]\u001b[A\n",
      " 84%|████████████████████████████████▉      | 1193/1415 [12:43<02:21,  1.57it/s]\u001b[A\n",
      " 84%|████████████████████████████████▉      | 1194/1415 [12:44<02:17,  1.60it/s]\u001b[A\n",
      " 84%|████████████████████████████████▉      | 1195/1415 [12:44<02:20,  1.56it/s]\u001b[A\n",
      " 85%|████████████████████████████████▉      | 1196/1415 [12:45<02:21,  1.55it/s]\u001b[A\n",
      " 85%|████████████████████████████████▉      | 1197/1415 [12:46<02:18,  1.58it/s]\u001b[A\n",
      " 85%|█████████████████████████████████      | 1198/1415 [12:46<02:17,  1.58it/s]\u001b[A\n",
      " 85%|█████████████████████████████████      | 1199/1415 [12:47<02:14,  1.60it/s]\u001b[A\n",
      " 85%|█████████████████████████████████      | 1200/1415 [12:48<02:14,  1.60it/s]\u001b[A\n",
      " 85%|█████████████████████████████████      | 1201/1415 [12:48<02:12,  1.62it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▏     | 1202/1415 [12:49<02:12,  1.61it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▏     | 1203/1415 [12:49<02:10,  1.62it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▏     | 1204/1415 [12:50<02:13,  1.59it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▏     | 1205/1415 [12:51<02:14,  1.57it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▏     | 1206/1415 [12:51<02:11,  1.59it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▎     | 1207/1415 [12:52<02:11,  1.59it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▎     | 1208/1415 [12:53<02:08,  1.61it/s]\u001b[A\n",
      " 85%|█████████████████████████████████▎     | 1209/1415 [12:53<02:09,  1.59it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▎     | 1210/1415 [12:54<02:07,  1.61it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▍     | 1211/1415 [12:55<02:10,  1.56it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▍     | 1212/1415 [12:55<02:10,  1.56it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▍     | 1213/1415 [12:56<02:11,  1.53it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▍     | 1214/1415 [12:57<02:12,  1.51it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▍     | 1215/1415 [12:57<02:09,  1.54it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▌     | 1216/1415 [12:58<02:08,  1.55it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▌     | 1217/1415 [12:58<02:06,  1.56it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▌     | 1218/1415 [12:59<02:06,  1.56it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▌     | 1219/1415 [13:00<02:04,  1.57it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▋     | 1220/1415 [13:00<02:04,  1.57it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▋     | 1221/1415 [13:01<02:01,  1.60it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▋     | 1222/1415 [13:02<02:01,  1.59it/s]\u001b[A\n",
      " 86%|█████████████████████████████████▋     | 1223/1415 [13:02<02:01,  1.58it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▋     | 1224/1415 [13:03<01:59,  1.60it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▊     | 1225/1415 [13:03<02:00,  1.58it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▊     | 1226/1415 [13:04<01:57,  1.60it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▊     | 1227/1415 [13:05<02:00,  1.56it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▊     | 1228/1415 [13:05<01:58,  1.58it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▊     | 1229/1415 [13:06<02:00,  1.54it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▉     | 1230/1415 [13:07<01:59,  1.55it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▉     | 1231/1415 [13:07<01:58,  1.55it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▉     | 1232/1415 [13:08<02:00,  1.52it/s]\u001b[A\n",
      " 87%|█████████████████████████████████▉     | 1233/1415 [13:09<01:56,  1.56it/s]\u001b[A\n",
      " 87%|██████████████████████████████████     | 1234/1415 [13:09<01:56,  1.56it/s]\u001b[A\n",
      " 87%|██████████████████████████████████     | 1235/1415 [13:10<01:54,  1.57it/s]\u001b[A\n",
      " 87%|██████████████████████████████████     | 1236/1415 [13:11<01:55,  1.55it/s]\u001b[A\n",
      " 87%|██████████████████████████████████     | 1237/1415 [13:11<01:52,  1.58it/s]\u001b[A\n",
      " 87%|██████████████████████████████████     | 1238/1415 [13:12<01:53,  1.57it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▏    | 1239/1415 [13:12<01:50,  1.59it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▏    | 1240/1415 [13:13<01:50,  1.59it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▏    | 1241/1415 [13:14<01:49,  1.59it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▏    | 1242/1415 [13:14<01:47,  1.60it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▎    | 1243/1415 [13:15<01:47,  1.60it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▎    | 1244/1415 [13:16<01:45,  1.61it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▎    | 1245/1415 [13:16<01:45,  1.61it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▎    | 1246/1415 [13:17<01:43,  1.63it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▎    | 1247/1415 [13:17<01:43,  1.62it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▍    | 1248/1415 [13:18<01:42,  1.63it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▍    | 1249/1415 [13:19<01:42,  1.62it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▍    | 1250/1415 [13:19<01:44,  1.58it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▍    | 1251/1415 [13:20<01:42,  1.60it/s]\u001b[A\n",
      " 88%|██████████████████████████████████▌    | 1252/1415 [13:21<01:42,  1.59it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▌    | 1253/1415 [13:21<01:40,  1.60it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▌    | 1254/1415 [13:22<01:43,  1.56it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▌    | 1255/1415 [13:22<01:41,  1.58it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▌    | 1256/1415 [13:23<01:41,  1.56it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▋    | 1257/1415 [13:24<01:39,  1.59it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▋    | 1258/1415 [13:24<01:39,  1.58it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▋    | 1259/1415 [13:25<01:39,  1.56it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▋    | 1260/1415 [13:26<01:37,  1.59it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▊    | 1261/1415 [13:26<01:38,  1.56it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▊    | 1262/1415 [13:27<01:36,  1.59it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▊    | 1263/1415 [13:28<01:36,  1.58it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▊    | 1264/1415 [13:28<01:35,  1.58it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▊    | 1265/1415 [13:29<01:38,  1.53it/s]\u001b[A\n",
      " 89%|██████████████████████████████████▉    | 1266/1415 [13:29<01:35,  1.57it/s]\u001b[A\n",
      " 90%|██████████████████████████████████▉    | 1267/1415 [13:30<01:35,  1.55it/s]\u001b[A\n",
      " 90%|██████████████████████████████████▉    | 1268/1415 [13:31<01:35,  1.54it/s]\u001b[A\n",
      " 90%|██████████████████████████████████▉    | 1269/1415 [13:31<01:33,  1.56it/s]\u001b[A\n",
      " 90%|███████████████████████████████████    | 1270/1415 [13:32<01:32,  1.56it/s]\u001b[A\n",
      " 90%|███████████████████████████████████    | 1271/1415 [13:33<01:31,  1.57it/s]\u001b[A\n",
      " 90%|███████████████████████████████████    | 1272/1415 [13:33<01:32,  1.55it/s]\u001b[A\n",
      " 90%|███████████████████████████████████    | 1273/1415 [13:34<01:29,  1.58it/s]\u001b[A\n",
      " 90%|███████████████████████████████████    | 1274/1415 [13:35<01:31,  1.55it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▏   | 1275/1415 [13:35<01:28,  1.58it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▏   | 1276/1415 [13:36<01:27,  1.58it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▏   | 1277/1415 [13:36<01:28,  1.56it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▏   | 1278/1415 [13:37<01:26,  1.59it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▎   | 1279/1415 [13:38<01:28,  1.54it/s]\u001b[A\n",
      " 90%|███████████████████████████████████▎   | 1280/1415 [13:38<01:25,  1.57it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▎   | 1281/1415 [13:39<01:25,  1.57it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▎   | 1282/1415 [13:40<01:23,  1.59it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▎   | 1283/1415 [13:40<01:23,  1.58it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▍   | 1284/1415 [13:41<01:23,  1.57it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▍   | 1285/1415 [13:42<01:22,  1.58it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▍   | 1286/1415 [13:42<01:22,  1.56it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▍   | 1287/1415 [13:43<01:20,  1.58it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▍   | 1288/1415 [13:43<01:20,  1.58it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▌   | 1289/1415 [13:44<01:18,  1.60it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▌   | 1290/1415 [13:45<01:18,  1.59it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▌   | 1291/1415 [13:45<01:16,  1.61it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▌   | 1292/1415 [13:46<01:16,  1.60it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▋   | 1293/1415 [13:47<01:15,  1.62it/s]\u001b[A\n",
      " 91%|███████████████████████████████████▋   | 1294/1415 [13:47<01:16,  1.59it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▋   | 1295/1415 [13:48<01:17,  1.55it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▋   | 1296/1415 [13:48<01:15,  1.57it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▋   | 1297/1415 [13:49<01:14,  1.58it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▊   | 1298/1415 [13:50<01:13,  1.60it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▊   | 1299/1415 [13:50<01:14,  1.55it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▊   | 1300/1415 [13:51<01:14,  1.54it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▊   | 1301/1415 [13:52<01:13,  1.56it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▉   | 1302/1415 [13:52<01:11,  1.59it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▉   | 1303/1415 [13:53<01:10,  1.59it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▉   | 1304/1415 [13:54<01:11,  1.56it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▉   | 1305/1415 [13:54<01:09,  1.59it/s]\u001b[A\n",
      " 92%|███████████████████████████████████▉   | 1306/1415 [13:55<01:08,  1.58it/s]\u001b[A\n",
      " 92%|████████████████████████████████████   | 1307/1415 [13:55<01:07,  1.61it/s]\u001b[A\n",
      " 92%|████████████████████████████████████   | 1308/1415 [13:56<01:06,  1.60it/s]\u001b[A\n",
      " 93%|████████████████████████████████████   | 1309/1415 [13:57<01:06,  1.60it/s]\u001b[A\n",
      " 93%|████████████████████████████████████   | 1310/1415 [13:57<01:05,  1.59it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▏  | 1311/1415 [13:58<01:04,  1.62it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▏  | 1312/1415 [13:59<01:03,  1.61it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▏  | 1313/1415 [13:59<01:04,  1.58it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▏  | 1314/1415 [14:00<01:02,  1.61it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▏  | 1315/1415 [14:00<01:02,  1.59it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▎  | 1316/1415 [14:01<01:02,  1.59it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▎  | 1317/1415 [14:02<01:02,  1.58it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▎  | 1318/1415 [14:02<01:00,  1.60it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▎  | 1319/1415 [14:03<01:00,  1.60it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▍  | 1320/1415 [14:04<00:59,  1.61it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▍  | 1321/1415 [14:04<00:58,  1.60it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▍  | 1322/1415 [14:05<00:59,  1.56it/s]\u001b[A\n",
      " 93%|████████████████████████████████████▍  | 1323/1415 [14:06<00:58,  1.57it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▍  | 1324/1415 [14:06<00:58,  1.57it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▌  | 1325/1415 [14:07<00:56,  1.60it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▌  | 1326/1415 [14:07<00:55,  1.59it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▌  | 1327/1415 [14:08<00:54,  1.62it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▌  | 1328/1415 [14:09<00:54,  1.61it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▋  | 1329/1415 [14:09<00:53,  1.62it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▋  | 1330/1415 [14:10<00:54,  1.57it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▋  | 1331/1415 [14:11<00:54,  1.54it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▋  | 1332/1415 [14:11<00:53,  1.55it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▋  | 1333/1415 [14:12<00:52,  1.56it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▊  | 1334/1415 [14:12<00:51,  1.58it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▊  | 1335/1415 [14:13<00:50,  1.58it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▊  | 1336/1415 [14:14<00:49,  1.61it/s]\u001b[A\n",
      " 94%|████████████████████████████████████▊  | 1337/1415 [14:14<00:49,  1.59it/s]\u001b[A\n",
      " 95%|████████████████████████████████████▉  | 1338/1415 [14:15<00:48,  1.58it/s]\u001b[A\n",
      " 95%|████████████████████████████████████▉  | 1339/1415 [14:16<00:49,  1.55it/s]\u001b[A\n",
      " 95%|████████████████████████████████████▉  | 1340/1415 [14:16<00:48,  1.53it/s]\u001b[A\n",
      " 95%|████████████████████████████████████▉  | 1341/1415 [14:17<00:47,  1.57it/s]\u001b[A\n",
      " 95%|████████████████████████████████████▉  | 1342/1415 [14:18<00:47,  1.55it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████  | 1343/1415 [14:18<00:46,  1.55it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████  | 1344/1415 [14:19<00:45,  1.56it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████  | 1345/1415 [14:19<00:44,  1.56it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████  | 1346/1415 [14:20<00:44,  1.57it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████▏ | 1347/1415 [14:21<00:42,  1.60it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████▏ | 1348/1415 [14:21<00:41,  1.60it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████▏ | 1349/1415 [14:22<00:41,  1.59it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████▏ | 1350/1415 [14:23<00:40,  1.61it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████▏ | 1351/1415 [14:23<00:39,  1.60it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▎ | 1352/1415 [14:24<00:38,  1.62it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▎ | 1353/1415 [14:24<00:38,  1.61it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▎ | 1354/1415 [14:25<00:37,  1.63it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▎ | 1355/1415 [14:26<00:37,  1.62it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▎ | 1356/1415 [14:26<00:36,  1.60it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 96%|█████████████████████████████████████▍ | 1357/1415 [14:27<00:36,  1.59it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▍ | 1358/1415 [14:28<00:36,  1.55it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▍ | 1359/1415 [14:28<00:35,  1.57it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▍ | 1360/1415 [14:29<00:35,  1.57it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▌ | 1361/1415 [14:30<00:34,  1.58it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▌ | 1362/1415 [14:30<00:34,  1.55it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▌ | 1363/1415 [14:31<00:33,  1.57it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▌ | 1364/1415 [14:31<00:32,  1.56it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████▌ | 1365/1415 [14:32<00:31,  1.58it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▋ | 1366/1415 [14:33<00:31,  1.53it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▋ | 1367/1415 [14:33<00:31,  1.54it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▋ | 1368/1415 [14:34<00:30,  1.55it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▋ | 1369/1415 [14:35<00:29,  1.55it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▊ | 1370/1415 [14:35<00:28,  1.57it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▊ | 1371/1415 [14:36<00:27,  1.58it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▊ | 1372/1415 [14:37<00:26,  1.59it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▊ | 1373/1415 [14:37<00:26,  1.58it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▊ | 1374/1415 [14:38<00:25,  1.61it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▉ | 1375/1415 [14:38<00:25,  1.59it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▉ | 1376/1415 [14:39<00:24,  1.59it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▉ | 1377/1415 [14:40<00:23,  1.60it/s]\u001b[A\n",
      " 97%|█████████████████████████████████████▉ | 1378/1415 [14:40<00:23,  1.58it/s]\u001b[A\n",
      " 97%|██████████████████████████████████████ | 1379/1415 [14:41<00:22,  1.58it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████ | 1380/1415 [14:42<00:22,  1.56it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████ | 1381/1415 [14:42<00:21,  1.59it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████ | 1382/1415 [14:43<00:20,  1.58it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████ | 1383/1415 [14:43<00:19,  1.60it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▏| 1384/1415 [14:44<00:19,  1.59it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▏| 1385/1415 [14:45<00:19,  1.55it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▏| 1386/1415 [14:45<00:18,  1.58it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▏| 1387/1415 [14:46<00:17,  1.56it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▎| 1388/1415 [14:47<00:17,  1.55it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▎| 1389/1415 [14:47<00:16,  1.54it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▎| 1390/1415 [14:48<00:16,  1.54it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▎| 1391/1415 [14:49<00:15,  1.51it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▎| 1392/1415 [14:49<00:14,  1.54it/s]\u001b[A\n",
      " 98%|██████████████████████████████████████▍| 1393/1415 [14:50<00:14,  1.54it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▍| 1394/1415 [14:51<00:13,  1.51it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▍| 1395/1415 [14:51<00:13,  1.51it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▍| 1396/1415 [14:52<00:12,  1.51it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▌| 1397/1415 [14:53<00:11,  1.52it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▌| 1398/1415 [14:53<00:11,  1.50it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▌| 1399/1415 [14:54<00:10,  1.55it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▌| 1400/1415 [14:55<00:09,  1.53it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▌| 1401/1415 [14:55<00:09,  1.53it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▋| 1402/1415 [14:56<00:08,  1.55it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▋| 1403/1415 [14:57<00:07,  1.55it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▋| 1404/1415 [14:57<00:07,  1.57it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▋| 1405/1415 [14:58<00:06,  1.57it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▊| 1406/1415 [14:58<00:05,  1.59it/s]\u001b[A\n",
      " 99%|██████████████████████████████████████▊| 1407/1415 [14:59<00:05,  1.56it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▊| 1408/1415 [15:00<00:04,  1.56it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▊| 1409/1415 [15:00<00:03,  1.53it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▊| 1410/1415 [15:01<00:03,  1.53it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▉| 1411/1415 [15:02<00:02,  1.51it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▉| 1412/1415 [15:02<00:02,  1.50it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▉| 1413/1415 [15:03<00:01,  1.51it/s]\u001b[A\n",
      "100%|██████████████████████████████████████▉| 1414/1415 [15:04<00:00,  1.52it/s]\u001b[A\n",
      "100%|███████████████████████████████████████| 1415/1415 [15:04<00:00,  1.56it/s]\u001b[A\n",
      "100%|████████████████████████████████████████████| 1/1 [15:14<00:00, 914.36s/it]\n"
     ]
    }
   ],
   "source": [
    "!satproc_extract_chips     \\\n",
    "    ./data_bariloche/img/NBR_img/NBR_2021-03-07_2021-03-29_0.tif \\\n",
    "    -o ./data_train_bariloche/onedate_FW/160_80/ \\\n",
    "    --size 160 \\\n",
    "    --step-size 80 \\\n",
    "    --aoi ./data/shp/gt/burn-area.geojson \\\n",
    "    --labels ./data/shp/gt/burn-area.geojson \\\n",
    "    --label-property 'class' \\\n",
    "    --classes 'fire' 'water' \\\n",
    "    --rescale \\\n",
    "    --rescale-mode percentiles \\\n",
    "    --upper-cut 98 --lower-cut 2\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### The options:\n",
    "\n",
    "* **the argument** the path to the source image or a folder with contains a set of images\n",
    "\n",
    "* **o** path to destiny images and masks\n",
    "* **size**size of the generated images and masks \n",
    "* **step-size** the size of the step in the slide windows process. if the step size is smaller than the size then an overlap is generated. This could be useful to increase the amount of training data.\n",
    "* **crs** it set an epsg in case the images don't have it. \n",
    "* **label-property** the name of the field in the shapefile where each category is defined. \n",
    "* **classes** name of the different categories (all split by a space)\n",
    "* **aoi** path to the shapefile that delimits the region of interest. It can be the same that the shapefile with the categories or other.\n",
    "* **rescale** to rescale the images to a range 1-255\n",
    "* **rescale-mode** Rescaled mode:  percentiles, custom_rgb, values. \n",
    "\n",
    "Percentiles mode will rescale each band if the image to a scale defined by the percentiles considering the upper and lower cuts. For example --upper-cut 98 --lower-cut 2\n",
    "\n",
    "The custom_rgb will rescale the 3 first bands from 0 to 0.3 and the rest of them considering the percentiles.\n",
    "\n",
    "The values mode will rescale all the bands in the image considering the min and max values defined. For example --Min 0 --Max 0.3   \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we should import the libraries needed "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install unetseg "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/ro/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/opt/miniconda3/envs/tf-gpu/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from unetseg.train import TrainConfig, train\n",
    "from unetseg.evaluate import plot_data_generator\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To set the parameters of the training process\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_config = TrainConfig(width=160,  #  size of the image (160, 320,etc)\n",
    "                     height=160,\n",
    "                     n_channels=1,  #  number of channels of the image rgb -> 3\n",
    "                     n_classes=2, # number of classes to clasify\n",
    "                     apply_image_augmentation=True, #  increase the amount of training data with a image augmentation process\n",
    "                     seed=42,\n",
    "                     epochs=15, \n",
    "                     batch_size=16, \n",
    "                     steps_per_epoch=100, \n",
    "                     early_stopping_patience=3, \n",
    "                     validation_split=0.2, \n",
    "                     test_split = 0.1,\n",
    "                     model_architecture = \"unet\",\n",
    "                     images_path=os.path.join('./data_train_bariloche','onedate_FW', '160_80'), #ruta a las imágenes\n",
    "                     model_path=os.path.join('./data/weights/', 'UNet_fire_160_80_spe100_fire_water.h5'),#  ruta al archivo de entrenamiento\n",
    "                     evaluate=True,\n",
    "                     class_weights=[0.6,0.4]) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To plot some exaples of images and masks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_data_generator(train_config=train_config, num_samples=1, fig_size=(10,5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/ro/.local/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-10-27 15:10:34.766540: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA\n",
      "2021-10-27 15:10:34.788832: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2299995000 Hz\n",
      "2021-10-27 15:10:34.789954: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5652f11688c0 executing computations on platform Host. Devices:\n",
      "2021-10-27 15:10:34.789980: I tensorflow/compiler/xla/service/service.cc:175]   StreamExecutor device (0): <undefined>, <undefined>\n",
      "2021-10-27 15:10:34.813096: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcuda.so.1\n",
      "2021-10-27 15:10:34.867513: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2021-10-27 15:10:34.868215: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1640] Found device 0 with properties: \n",
      "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
      "pciBusID: 0000:00:04.0\n",
      "2021-10-27 15:10:34.868458: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcudart.so.10.0'; dlerror: libcudart.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:34.868630: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcublas.so.10.0'; dlerror: libcublas.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:34.868791: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcufft.so.10.0'; dlerror: libcufft.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:34.868935: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcurand.so.10.0'; dlerror: libcurand.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:34.869071: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcusolver.so.10.0'; dlerror: libcusolver.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:34.869211: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcusparse.so.10.0'; dlerror: libcusparse.so.10.0: cannot open shared object file: No such file or directory\n",
      "2021-10-27 15:10:35.001768: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.7\n",
      "2021-10-27 15:10:35.001830: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1663] Cannot dlopen some GPU libraries. Skipping registering GPU devices...\n",
      "2021-10-27 15:10:35.221158: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1181] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
      "2021-10-27 15:10:35.221190: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1187]      0 \n",
      "2021-10-27 15:10:35.221196: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1200] 0:   N \n",
      "2021-10-27 15:10:35.224125: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2021-10-27 15:10:35.224912: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5652f58b76c0 executing computations on platform CUDA. Devices:\n",
      "2021-10-27 15:10:35.224938: I tensorflow/compiler/xla/service/service.cc:175]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
      "2021-10-27 15:10:35.269182: W tensorflow/compiler/jit/mark_for_compilation_pass.cc:1412] (One-time warning): Not using XLA:CPU for cluster because envvar TF_XLA_FLAGS=--tf_xla_cpu_global_jit was not set.  If you want XLA:CPU, either set that envvar, or use experimental_jit_scope to enable XLA:CPU.  To confirm that XLA is active, pass --vmodule=xla_compilation_cache=1 (as a proper command-line flag, not via TF_XLA_FLAGS) or set the envvar XLA_FLAGS=--xla_hlo_profile.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/ro/unetseg/src/unetseg/train.py:425: The name tf.metrics.mean_iou is deprecated. Please use tf.compat.v1.metrics.mean_iou instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/ro/.local/lib/python3.7/site-packages/tensorflow/python/ops/metrics_impl.py:1178: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "WARNING:tensorflow:From /home/ro/.local/lib/python3.7/site-packages/tensorflow/python/ops/metrics_impl.py:1179: div (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Deprecated in favor of operator or tf.math.divide.\n",
      "WARNING:tensorflow:From /home/ro/unetseg/src/unetseg/train.py:426: The name tf.local_variables_initializer is deprecated. Please use tf.compat.v1.local_variables_initializer instead.\n",
      "\n",
      "Model: \"model_1\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 160, 160, 1)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 160, 160, 32) 320         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 160, 160, 32) 9248        conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, 80, 80, 32)   0           conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 80, 80, 32)   128         max_pooling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 80, 80, 64)   18496       batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 80, 80, 64)   36928       conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)  (None, 40, 40, 64)   0           conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 40, 40, 64)   0           max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 40, 40, 64)   256         dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 40, 40, 128)  73856       batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 40, 40, 128)  147584      conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)  (None, 20, 20, 128)  0           conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 20, 20, 128)  0           max_pooling2d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 20, 20, 128)  512         dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 20, 20, 256)  295168      batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 20, 20, 256)  590080      conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2D)  (None, 10, 10, 256)  0           conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_3 (Dropout)             (None, 10, 10, 256)  0           max_pooling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 10, 10, 256)  1024        dropout_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_9 (Conv2D)               (None, 10, 10, 512)  1180160     batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_10 (Conv2D)              (None, 10, 10, 512)  2359808     conv2d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2D)  (None, 5, 5, 512)    0           conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_4 (Dropout)             (None, 5, 5, 512)    0           max_pooling2d_5[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_11 (Conv2D)              (None, 5, 5, 1024)   4719616     dropout_4[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_12 (Conv2D)              (None, 5, 5, 1024)   9438208     conv2d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_1 (Conv2DTrans (None, 10, 10, 512)  2097664     conv2d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 10, 10, 1024) 0           conv2d_transpose_1[0][0]         \n",
      "                                                                 conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 10, 10, 1024) 4096        concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_13 (Conv2D)              (None, 10, 10, 512)  4719104     batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_14 (Conv2D)              (None, 10, 10, 512)  2359808     conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_5 (Dropout)             (None, 10, 10, 512)  0           conv2d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_2 (Conv2DTrans (None, 20, 20, 256)  524544      dropout_5[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 20, 20, 512)  0           conv2d_transpose_2[0][0]         \n",
      "                                                                 conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 20, 20, 512)  2048        concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_15 (Conv2D)              (None, 20, 20, 256)  1179904     batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_16 (Conv2D)              (None, 20, 20, 256)  590080      conv2d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_6 (Dropout)             (None, 20, 20, 256)  0           conv2d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_3 (Conv2DTrans (None, 40, 40, 128)  131200      dropout_6[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 40, 40, 256)  0           conv2d_transpose_3[0][0]         \n",
      "                                                                 conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 40, 40, 256)  1024        concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_17 (Conv2D)              (None, 40, 40, 128)  295040      batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_18 (Conv2D)              (None, 40, 40, 128)  147584      conv2d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_7 (Dropout)             (None, 40, 40, 128)  0           conv2d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_4 (Conv2DTrans (None, 80, 80, 64)   32832       dropout_7[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 80, 80, 128)  0           conv2d_transpose_4[0][0]         \n",
      "                                                                 conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 80, 80, 128)  512         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_19 (Conv2D)              (None, 80, 80, 64)   73792       batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_20 (Conv2D)              (None, 80, 80, 64)   36928       conv2d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_8 (Dropout)             (None, 80, 80, 64)   0           conv2d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_5 (Conv2DTrans (None, 160, 160, 32) 8224        dropout_8[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_5 (Concatenate)     (None, 160, 160, 64) 0           conv2d_transpose_5[0][0]         \n",
      "                                                                 conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_21 (Conv2D)              (None, 160, 160, 32) 18464       concatenate_5[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_22 (Conv2D)              (None, 160, 160, 32) 9248        conv2d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_23 (Conv2D)              (None, 160, 160, 2)  66          conv2d_22[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 31,103,554\n",
      "Trainable params: 31,098,754\n",
      "Non-trainable params: 4,800\n",
      "__________________________________________________________________________________________________\n",
      "None\n",
      "All images: 1415\n",
      "Num. training images: 1018\n",
      "Num. validation images: 255\n",
      "Num. test images: 142\n",
      "Compile and fit the UNet model\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/ro/.local/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n",
      "Epoch 1/15\n",
      "100/100 [==============================] - 748s 7s/step - loss: 0.3822 - mean_iou: 0.4545 - val_loss: 1.4134 - val_mean_iou: 0.5138\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 1.41337, saving model to ./data/weights/UNet_fire_160_80_spe100_fire_water.h5\n",
      "Epoch 2/15\n",
      "100/100 [==============================] - 740s 7s/step - loss: 0.2586 - mean_iou: 0.5375 - val_loss: 15.0372 - val_mean_iou: 0.5385\n",
      "\n",
      "Epoch 00002: val_loss did not improve from 1.41337\n",
      "Epoch 3/15\n",
      "100/100 [==============================] - 742s 7s/step - loss: 0.2275 - mean_iou: 0.5406 - val_loss: 5.0788 - val_mean_iou: 0.5419\n",
      "\n",
      "Epoch 00003: val_loss did not improve from 1.41337\n",
      "Epoch 4/15\n",
      "100/100 [==============================] - 726s 7s/step - loss: 0.2148 - mean_iou: 0.5438 - val_loss: 0.1401 - val_mean_iou: 0.5544\n",
      "\n",
      "Epoch 00004: val_loss improved from 1.41337 to 0.14010, saving model to ./data/weights/UNet_fire_160_80_spe100_fire_water.h5\n",
      "Epoch 5/15\n",
      "100/100 [==============================] - 755s 8s/step - loss: 0.2196 - mean_iou: 0.5609 - val_loss: 0.2283 - val_mean_iou: 0.5681\n",
      "\n",
      "Epoch 00005: val_loss did not improve from 0.14010\n",
      "Epoch 6/15\n",
      "100/100 [==============================] - 727s 7s/step - loss: 0.1956 - mean_iou: 0.5759 - val_loss: 0.1723 - val_mean_iou: 0.5819\n",
      "\n",
      "Epoch 00006: val_loss did not improve from 0.14010\n",
      "Epoch 7/15\n",
      "100/100 [==============================] - 735s 7s/step - loss: 0.2056 - mean_iou: 0.5846 - val_loss: 0.1869 - val_mean_iou: 0.5894\n",
      "\n",
      "Epoch 00007: val_loss did not improve from 0.14010\n",
      "Epoch 00007: early stopping\n",
      "**** Final  metrics ****\n",
      "Loss: 0.19590342044830322\n",
      "Mean IoU: 0.5909033417701721\n"
     ]
    }
   ],
   "source": [
    "res_config = train(train_config)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Print some metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(16,4))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.plot(res_config.history['loss'])\n",
    "plt.plot(res_config.history['val_loss'])\n",
    "plt.title('Loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Val'], loc='upper left')\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(res_config.history['mean_iou'])\n",
    "plt.plot(res_config.history['val_mean_iou'])\n",
    "plt.title('mean_iou')\n",
    "plt.ylabel('val_mean_iou')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Val'], loc='upper left')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyNKlws8+6PcuvJrSWdSmTxJ",
   "include_colab_link": true,
   "name": "2 - Entrenamiento",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.11"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}