src/codebase/iPython/Cumulative_performance/AWA2-VIT.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 1;\n",
       "                var nbb_unformatted_code = \"%load_ext nb_black\\n\\nimport os\\nimport sys\\n\\nsys.path.append(\\n    os.path.abspath(\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/codebase\\\")\\n)\\nimport torch.backends.cudnn as cudnn\\nimport random\\nimport time\\nfrom collections import OrderedDict\\nimport pickle\\nimport numpy as np\\nimport pandas as pd\\nimport torch\\nimport torch.nn.functional as F\\nfrom torch.utils.data import DataLoader\\nfrom tqdm import tqdm\\nfrom torch.nn.functional import one_hot\\nimport sklearn.metrics as metrics\\nimport utils\\nfrom Explainer.loss_F import loss_fn_kd, entropy_loss\\nfrom Explainer.models.Gated_Logic_Net import Gated_Logic_Net\\nfrom Explainer.models.explainer import Explainer\\nfrom Explainer.models.pi import Pi\\nfrom dataset.dataset_cubs import Dataset_cub_for_explainer\\nfrom dataset.utils_dataset import get_dataset_with_image_and_attributes\\nfrom Explainer.loss_F import loss_fn_kd, entropy_loss, Selective_Distillation_Loss\\nfrom Explainer.models.concepts import Conceptizator\\nimport Plots.plot_for_paper as plot\\nimport matplotlib.pyplot as plt\";\n",
       "                var nbb_formatted_code = \"%load_ext nb_black\\n\\nimport os\\nimport sys\\n\\nsys.path.append(\\n    os.path.abspath(\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/codebase\\\")\\n)\\nimport torch.backends.cudnn as cudnn\\nimport random\\nimport time\\nfrom collections import OrderedDict\\nimport pickle\\nimport numpy as np\\nimport pandas as pd\\nimport torch\\nimport torch.nn.functional as F\\nfrom torch.utils.data import DataLoader\\nfrom tqdm import tqdm\\nfrom torch.nn.functional import one_hot\\nimport sklearn.metrics as metrics\\nimport utils\\nfrom Explainer.loss_F import loss_fn_kd, entropy_loss\\nfrom Explainer.models.Gated_Logic_Net import Gated_Logic_Net\\nfrom Explainer.models.explainer import Explainer\\nfrom Explainer.models.pi import Pi\\nfrom dataset.dataset_cubs import Dataset_cub_for_explainer\\nfrom dataset.utils_dataset import get_dataset_with_image_and_attributes\\nfrom Explainer.loss_F import loss_fn_kd, entropy_loss, Selective_Distillation_Loss\\nfrom Explainer.models.concepts import Conceptizator\\nimport Plots.plot_for_paper as plot\\nimport matplotlib.pyplot as plt\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%load_ext nb_black\n",
    "\n",
    "import os\n",
    "import sys\n",
    "\n",
    "sys.path.append(\n",
    "    os.path.abspath(\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/codebase\")\n",
    ")\n",
    "import torch.backends.cudnn as cudnn\n",
    "import random\n",
    "import time\n",
    "from collections import OrderedDict\n",
    "import pickle\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import torch\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import DataLoader\n",
    "from tqdm import tqdm\n",
    "from torch.nn.functional import one_hot\n",
    "import sklearn.metrics as metrics\n",
    "import utils\n",
    "from Explainer.loss_F import loss_fn_kd, entropy_loss\n",
    "from Explainer.models.Gated_Logic_Net import Gated_Logic_Net\n",
    "from Explainer.models.explainer import Explainer\n",
    "from Explainer.models.pi import Pi\n",
    "from dataset.dataset_cubs import Dataset_cub_for_explainer\n",
    "from dataset.utils_dataset import get_dataset_with_image_and_attributes\n",
    "from Explainer.loss_F import loss_fn_kd, entropy_loss, Selective_Distillation_Loss\n",
    "from Explainer.models.concepts import Conceptizator\n",
    "import Plots.plot_for_paper as plot\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Device: cuda\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 2;\n",
       "                var nbb_unformatted_code = \"device = utils.get_device()\\nprint(f\\\"Device: {device}\\\")\";\n",
       "                var nbb_formatted_code = \"device = utils.get_device()\\nprint(f\\\"Device: {device}\\\")\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "device = utils.get_device()\n",
    "print(f\"Device: {device}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "[0.01]\n",
      "[0.2]\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([1436, 50])\n",
      "torch.Size([1436, 50])\n",
      "torch.Size([1436])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 3;\n",
       "                var nbb_unformatted_code = \"lr = 0.01\\ncov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"iter1\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr)\\nprint(args.cov)\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter1\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_1 = torch.load(\\n    os.path.join(\\n        root, experiment, iteration, expert_type, output, \\\"test_tensor_preds.pt\\\"\\n    )\\n)\\n\\ntest_tensor_preds_bb_1 = torch.load(\\n    os.path.join(\\n        root, experiment, iteration, expert_type, output, \\\"test_tensor_preds_bb.pt\\\"\\n    )\\n)\\n\\ntest_tensor_y_1 = torch.load(\\n    os.path.join(root, experiment, iteration, expert_type, output, \\\"test_tensor_y.pt\\\")\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_1.size())\\nprint(test_tensor_preds_bb_1.size())\\nprint(test_tensor_y_1.size())\";\n",
       "                var nbb_formatted_code = \"lr = 0.01\\ncov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"iter1\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr)\\nprint(args.cov)\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter1\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_1 = torch.load(\\n    os.path.join(\\n        root, experiment, iteration, expert_type, output, \\\"test_tensor_preds.pt\\\"\\n    )\\n)\\n\\ntest_tensor_preds_bb_1 = torch.load(\\n    os.path.join(\\n        root, experiment, iteration, expert_type, output, \\\"test_tensor_preds_bb.pt\\\"\\n    )\\n)\\n\\ntest_tensor_y_1 = torch.load(\\n    os.path.join(root, experiment, iteration, expert_type, output, \\\"test_tensor_y.pt\\\")\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_1.size())\\nprint(test_tensor_preds_bb_1.size())\\nprint(test_tensor_y_1.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lr = 0.01\n",
    "cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"iter1\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr)\n",
    "print(args.cov)\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter1\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_1 = torch.load(\n",
    "    os.path.join(\n",
    "        root, experiment, iteration, expert_type, output, \"test_tensor_preds.pt\"\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_1 = torch.load(\n",
    "    os.path.join(\n",
    "        root, experiment, iteration, expert_type, output, \"test_tensor_preds_bb.pt\"\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_1 = torch.load(\n",
    "    os.path.join(root, experiment, iteration, expert_type, output, \"test_tensor_y.pt\")\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_1.size())\n",
    "print(test_tensor_preds_bb_1.size())\n",
    "print(test_tensor_y_1.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "0.01\n",
      "0.2\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([1643, 50])\n",
      "torch.Size([1643, 50])\n",
      "torch.Size([1643])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 4;\n",
       "                var nbb_unformatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter2\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter2\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_2.size())\\nprint(test_tensor_preds_bb_2.size())\\nprint(test_tensor_y_2.size())\";\n",
       "                var nbb_formatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter2\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter2\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_2 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_2.size())\\nprint(test_tensor_preds_bb_2.size())\\nprint(test_tensor_y_2.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_lr = 0.01\n",
    "base_cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        \"iter2\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr[-1])\n",
    "print(args.cov[-1])\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter2\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_2 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_2 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_2 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_2.size())\n",
    "print(test_tensor_preds_bb_2.size())\n",
    "print(test_tensor_y_2.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "0.01\n",
      "0.2\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([1432, 50])\n",
      "torch.Size([1432, 50])\n",
      "torch.Size([1432])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 5;\n",
       "                var nbb_unformatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter3\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter3\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_3.size())\\nprint(test_tensor_preds_bb_3.size())\\nprint(test_tensor_y_3.size())\";\n",
       "                var nbb_formatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter3\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter3\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_3 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_3.size())\\nprint(test_tensor_preds_bb_3.size())\\nprint(test_tensor_y_3.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_lr = 0.01\n",
    "base_cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        \"iter3\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr[-1])\n",
    "print(args.cov[-1])\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter3\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_3 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_3 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_3 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_3.size())\n",
    "print(test_tensor_preds_bb_3.size())\n",
    "print(test_tensor_y_3.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "0.01\n",
      "0.2\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([1002, 50])\n",
      "torch.Size([1002, 50])\n",
      "torch.Size([1002])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 6;\n",
       "                var nbb_unformatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter4\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter4\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_4.size())\\nprint(test_tensor_preds_bb_4.size())\\nprint(test_tensor_y_4.size())\";\n",
       "                var nbb_formatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter4\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter4\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_4 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_4.size())\\nprint(test_tensor_preds_bb_4.size())\\nprint(test_tensor_y_4.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_lr = 0.01\n",
    "base_cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        \"iter4\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr[-1])\n",
    "print(args.cov[-1])\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter4\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_4 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_4 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_4 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_4.size())\n",
    "print(test_tensor_preds_bb_4.size())\n",
    "print(test_tensor_y_4.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "0.01\n",
      "0.2\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([918, 50])\n",
      "torch.Size([918, 50])\n",
      "torch.Size([918])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 7;\n",
       "                var nbb_unformatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter5\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter5\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_5.size())\\nprint(test_tensor_preds_bb_5.size())\\nprint(test_tensor_y_5.size())\";\n",
       "                var nbb_formatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter5\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter5\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_5 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_5.size())\\nprint(test_tensor_preds_bb_5.size())\\nprint(test_tensor_y_5.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_lr = 0.01\n",
    "base_cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        \"iter5\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr[-1])\n",
    "print(args.cov[-1])\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter5\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_5 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_5 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_5 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_5.size())\n",
    "print(test_tensor_preds_bb_5.size())\n",
    "print(test_tensor_y_5.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "########################\n",
      "0.01\n",
      "0.2\n",
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([655, 50])\n",
      "torch.Size([655, 50])\n",
      "torch.Size([655])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 8;\n",
       "                var nbb_unformatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter6\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter6\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_6.size())\\nprint(test_tensor_preds_bb_6.size())\\nprint(test_tensor_y_6.size())\";\n",
       "                var nbb_formatted_code = \"base_lr = 0.01\\nbase_cov = 0.2\\npickle_in = open(\\n    os.path.join(\\n        f\\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\\\",\\n        \\\"cov_0.2_lr_0.01\\\",\\n        \\\"iter6\\\",\\n        \\\"explainer\\\",\\n        \\\"test_explainer_configs.pkl\\\",\\n    ),\\n    \\\"rb\\\",\\n)\\nargs = pickle.load(pickle_in)\\nn_classes = len(args.labels)\\nx_to_bool = 0.5\\ntop_k_explanations = 50\\nconcept_names = args.concept_names\\nprint(\\\"########################\\\")\\nprint(args.lr[-1])\\nprint(args.cov[-1])\\n\\nuse_concepts_as_pi_input = True\\nexplainer_init = \\\"none\\\"\\n\\nexperiment_folder = (\\n    f\\\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\\\"\\n    f\\\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\\\"\\n    f\\\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\\\"\\n    f\\\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\\\"\\n    f\\\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\\\"\\n    f\\\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\\\"\\n)\\n\\nroot = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\\\"\\nexperiment = f\\\"explainer/ViT-B_16/{experiment_folder}\\\"\\niteration = \\\"iter6\\\"\\nexpert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\n\\ntest_tensor_preds_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_6 = torch.load(\\n    os.path.join(\\n        root,\\n        experiment,\\n        \\\"cov_0.2_lr_0.01\\\",\\n        iteration,\\n        expert_type,\\n        output,\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_6.size())\\nprint(test_tensor_preds_bb_6.size())\\nprint(test_tensor_y_6.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_lr = 0.01\n",
    "base_cov = 0.2\n",
    "pickle_in = open(\n",
    "    os.path.join(\n",
    "        f\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_{base_lr}_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_{base_cov}_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none\",\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        \"iter6\",\n",
    "        \"explainer\",\n",
    "        \"test_explainer_configs.pkl\",\n",
    "    ),\n",
    "    \"rb\",\n",
    ")\n",
    "args = pickle.load(pickle_in)\n",
    "n_classes = len(args.labels)\n",
    "x_to_bool = 0.5\n",
    "top_k_explanations = 50\n",
    "concept_names = args.concept_names\n",
    "print(\"########################\")\n",
    "print(args.lr[-1])\n",
    "print(args.cov[-1])\n",
    "\n",
    "use_concepts_as_pi_input = True\n",
    "explainer_init = \"none\"\n",
    "\n",
    "experiment_folder = (\n",
    "    f\"lr_{args.lr[0]}_epochs_{args.epochs}_temperature-lens_{args.temperature_lens}\"\n",
    "    f\"_use-concepts-as-pi-input_{use_concepts_as_pi_input}_input-size-pi_{args.input_size_pi}\"\n",
    "    f\"_cov_{args.cov[0]}_alpha_{args.alpha}_selection-threshold_{args.selection_threshold}\"\n",
    "    f\"_lambda-lens_{args.lambda_lens}_alpha-KD_{args.alpha_KD}\"\n",
    "    f\"_temperature-KD_{float(args.temperature_KD)}_hidden-layers_{len(args.hidden_nodes)}\"\n",
    "    f\"_layer_{args.layer}_explainer_init_{explainer_init if not args.explainer_init else args.explainer_init}\"\n",
    ")\n",
    "\n",
    "root = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/\"\n",
    "experiment = f\"explainer/ViT-B_16/{experiment_folder}\"\n",
    "iteration = \"iter6\"\n",
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "\n",
    "test_tensor_preds_6 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_6 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_6 = torch.load(\n",
    "    os.path.join(\n",
    "        root,\n",
    "        experiment,\n",
    "        \"cov_0.2_lr_0.01\",\n",
    "        iteration,\n",
    "        expert_type,\n",
    "        output,\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_6.size())\n",
    "print(test_tensor_preds_bb_6.size())\n",
    "print(test_tensor_y_6.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 9;\n",
       "                var nbb_unformatted_code = \"expert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\npreds = torch.cat((test_tensor_preds_1, test_tensor_preds_2, test_tensor_preds_3, \\n                   test_tensor_preds_4, test_tensor_preds_5, \\n                   test_tensor_preds_6\\n                   \\n                  ), dim=0)\\ngt = torch.cat((test_tensor_y_1, test_tensor_y_2, test_tensor_y_3 ,test_tensor_y_4, \\n                test_tensor_y_5, test_tensor_y_6\\n               ), dim=0)\\nbb = torch.cat((\\n    test_tensor_preds_bb_1.cpu(), test_tensor_preds_bb_2.cpu(), test_tensor_preds_bb_3.cpu(),\\n    test_tensor_preds_bb_4.cpu(), \\n    test_tensor_preds_bb_5.cpu(), test_tensor_preds_bb_6.cpu()\\n), \\n               dim=0)\";\n",
       "                var nbb_formatted_code = \"expert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\npreds = torch.cat(\\n    (\\n        test_tensor_preds_1,\\n        test_tensor_preds_2,\\n        test_tensor_preds_3,\\n        test_tensor_preds_4,\\n        test_tensor_preds_5,\\n        test_tensor_preds_6,\\n    ),\\n    dim=0,\\n)\\ngt = torch.cat(\\n    (\\n        test_tensor_y_1,\\n        test_tensor_y_2,\\n        test_tensor_y_3,\\n        test_tensor_y_4,\\n        test_tensor_y_5,\\n        test_tensor_y_6,\\n    ),\\n    dim=0,\\n)\\nbb = torch.cat(\\n    (\\n        test_tensor_preds_bb_1.cpu(),\\n        test_tensor_preds_bb_2.cpu(),\\n        test_tensor_preds_bb_3.cpu(),\\n        test_tensor_preds_bb_4.cpu(),\\n        test_tensor_preds_bb_5.cpu(),\\n        test_tensor_preds_bb_6.cpu(),\\n    ),\\n    dim=0,\\n)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "preds = torch.cat((test_tensor_preds_1, test_tensor_preds_2, test_tensor_preds_3, \n",
    "                   test_tensor_preds_4, test_tensor_preds_5, \n",
    "                   test_tensor_preds_6\n",
    "                   \n",
    "                  ), dim=0)\n",
    "gt = torch.cat((test_tensor_y_1, test_tensor_y_2, test_tensor_y_3 ,test_tensor_y_4, \n",
    "                test_tensor_y_5, test_tensor_y_6\n",
    "               ), dim=0)\n",
    "bb = torch.cat((\n",
    "    test_tensor_preds_bb_1.cpu(), test_tensor_preds_bb_2.cpu(), test_tensor_preds_bb_3.cpu(),\n",
    "    test_tensor_preds_bb_4.cpu(), \n",
    "    test_tensor_preds_bb_5.cpu(), test_tensor_preds_bb_6.cpu()\n",
    "), \n",
    "               dim=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "g_tot_acc: 97.57267852102738 (%)\n",
      "bb_tot_acc: 98.33474456675134 (%)\n",
      "g_tot_f1: 97.57267852102738 (%)\n",
      "g_tot_precision: 97.57267852102738 (%)\n",
      "g_tot_recall: 97.57267852102738 (%)\n",
      "bb_tot_recall: 98.33474456675134 (%)\n",
      "total samples covered by g: 7086 (out of 7465)\n",
      "total coverage by g: 0.9492297387809779\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 10;\n",
       "                var nbb_unformatted_code = \"preds_tot_np = preds.argmax(dim=1).numpy()\\ngt_tot_np = gt.numpy()\\nbb_tot_np = bb.argmax(dim=1).numpy()\\nbb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\\ng_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\\ng_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\nbb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\\\"micro\\\")\\ng_tot_cov = preds.size(0) / 7465\\n\\n\\nprint(f\\\"g_tot_acc: {g_tot_acc * 100} (%)\\\")\\nprint(f\\\"bb_tot_acc: {bb_tot_acc * 100} (%)\\\")\\nprint(f\\\"g_tot_f1: {g_tot_f1 * 100} (%)\\\")\\nprint(f\\\"g_tot_precision: {g_tot_precision * 100} (%)\\\")\\nprint(f\\\"g_tot_recall: {g_tot_recall * 100} (%)\\\")\\nprint(f\\\"bb_tot_recall: {bb_tot_recall * 100} (%)\\\")\\nprint(f\\\"total samples covered by g: {preds.size(0)} (out of {7465})\\\")\\nprint(f\\\"total coverage by g: {g_tot_cov}\\\")\";\n",
       "                var nbb_formatted_code = \"preds_tot_np = preds.argmax(dim=1).numpy()\\ngt_tot_np = gt.numpy()\\nbb_tot_np = bb.argmax(dim=1).numpy()\\nbb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\\ng_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\\ng_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\nbb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\\\"micro\\\")\\ng_tot_cov = preds.size(0) / 7465\\n\\n\\nprint(f\\\"g_tot_acc: {g_tot_acc * 100} (%)\\\")\\nprint(f\\\"bb_tot_acc: {bb_tot_acc * 100} (%)\\\")\\nprint(f\\\"g_tot_f1: {g_tot_f1 * 100} (%)\\\")\\nprint(f\\\"g_tot_precision: {g_tot_precision * 100} (%)\\\")\\nprint(f\\\"g_tot_recall: {g_tot_recall * 100} (%)\\\")\\nprint(f\\\"bb_tot_recall: {bb_tot_recall * 100} (%)\\\")\\nprint(f\\\"total samples covered by g: {preds.size(0)} (out of {7465})\\\")\\nprint(f\\\"total coverage by g: {g_tot_cov}\\\")\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "preds_tot_np = preds.argmax(dim=1).numpy()\n",
    "gt_tot_np = gt.numpy()\n",
    "bb_tot_np = bb.argmax(dim=1).numpy()\n",
    "bb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\n",
    "g_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\n",
    "g_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "g_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "g_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "bb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\"micro\")\n",
    "g_tot_cov = preds.size(0) / 7465\n",
    "\n",
    "\n",
    "print(f\"g_tot_acc: {g_tot_acc * 100} (%)\")\n",
    "print(f\"bb_tot_acc: {bb_tot_acc * 100} (%)\")\n",
    "print(f\"g_tot_f1: {g_tot_f1 * 100} (%)\")\n",
    "print(f\"g_tot_precision: {g_tot_precision * 100} (%)\")\n",
    "print(f\"g_tot_recall: {g_tot_recall * 100} (%)\")\n",
    "print(f\"bb_tot_recall: {bb_tot_recall * 100} (%)\")\n",
    "print(f\"total samples covered by g: {preds.size(0)} (out of {7465})\")\n",
    "print(f\"total coverage by g: {g_tot_cov}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      " << Test sizes >>\n",
      "torch.Size([379, 50])\n",
      "torch.Size([379, 50])\n",
      "torch.Size([379])\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 11;\n",
       "                var nbb_unformatted_code = \"test_tensor_preds_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_preds_residual.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_residual.size())\\nprint(test_tensor_preds_bb_residual.size())\\nprint(test_tensor_y_residual.size())\";\n",
       "                var nbb_formatted_code = \"test_tensor_preds_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_preds_residual.pt\\\",\\n    )\\n)\\n\\ntest_tensor_preds_bb_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_preds_bb.pt\\\",\\n    )\\n)\\n\\ntest_tensor_y_residual = torch.load(\\n    os.path.join(\\n        \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\\\",\\n        \\\"test_tensor_y.pt\\\",\\n    )\\n)\\n\\n\\nprint(\\\"\\\\n\\\\n << Test sizes >>\\\")\\nprint(test_tensor_preds_residual.size())\\nprint(test_tensor_preds_bb_residual.size())\\nprint(test_tensor_y_residual.size())\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_tensor_preds_residual = torch.load(\n",
    "    os.path.join(\n",
    "        \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\",\n",
    "        \"test_tensor_preds_residual.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_preds_bb_residual = torch.load(\n",
    "    os.path.join(\n",
    "        \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\",\n",
    "        \"test_tensor_preds_bb.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "test_tensor_y_residual = torch.load(\n",
    "    os.path.join(\n",
    "        \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/out/awa2/explainer/ViT-B_16/lr_0.01_epochs_500_temperature-lens_6.0_use-concepts-as-pi-input_True_input-size-pi_2048_cov_0.2_alpha_0.5_selection-threshold_0.5_lambda-lens_0.0001_alpha-KD_0.99_temperature-KD_10.0_hidden-layers_1_layer_VIT_explainer_init_none/cov_0.2_lr_0.01/iter6/bb/residual_outputs\",\n",
    "        \"test_tensor_y.pt\",\n",
    "    )\n",
    ")\n",
    "\n",
    "\n",
    "print(\"\\n\\n << Test sizes >>\")\n",
    "print(test_tensor_preds_residual.size())\n",
    "print(test_tensor_preds_bb_residual.size())\n",
    "print(test_tensor_y_residual.size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(0.9657)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 12;\n",
       "                var nbb_unformatted_code = \"torch.sum(\\n    test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual\\n) / test_tensor_preds_residual.size(0)\";\n",
       "                var nbb_formatted_code = \"torch.sum(\\n    test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual\\n) / test_tensor_preds_residual.size(0)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "torch.sum(\n",
    "    test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual\n",
    ") / test_tensor_preds_residual.size(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "g_tot_acc: 97.52176825184193 (%)\n",
      "bb_tot_acc: 98.24514400535834 (%)\n",
      "g_tot_f1: 97.52176825184193 (%)\n",
      "g_tot_precision: 97.52176825184193 (%)\n",
      "g_tot_recall: 97.52176825184193 (%)\n",
      "bb_tot_recall: 98.24514400535834 (%)\n",
      "total samples covered by g: 7465 (out of 7465)\n",
      "total coverage by g: 1.0\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 13;\n",
       "                var nbb_unformatted_code = \"expert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\npreds = torch.cat((test_tensor_preds_1, test_tensor_preds_2, test_tensor_preds_3, \\n                   test_tensor_preds_4, test_tensor_preds_5, test_tensor_preds_6,\\n                     test_tensor_preds_residual,\\n#                    test_tensor_preds_7, test_tensor_preds_8\\n                  ), dim=0)\\ngt = torch.cat((test_tensor_y_1, test_tensor_y_2, test_tensor_y_3 ,\\n                test_tensor_y_4, test_tensor_y_5, test_tensor_y_6,\\n                test_tensor_y_residual\\n#                test_tensor_y_7, test_tensor_y_8\\n               ), dim=0)\\nbb = torch.cat((\\n    test_tensor_preds_bb_1.cpu(), test_tensor_preds_bb_2.cpu(), test_tensor_preds_bb_3.cpu(),\\n    test_tensor_preds_bb_4.cpu(), test_tensor_preds_bb_5.cpu(), test_tensor_preds_bb_6.cpu(),\\n    test_tensor_preds_bb_residual.cpu()\\n#     test_tensor_preds_bb_7.cpu(), test_tensor_preds_bb_8.cpu()\\n), \\n               dim=0)\\n\\npreds_tot_np = preds.argmax(dim=1).numpy()\\ngt_tot_np = gt.numpy()\\nbb_tot_np = bb.argmax(dim=1).numpy()\\nbb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\\ng_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\\ng_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\nbb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\\\"micro\\\")\\ng_tot_cov = preds.size(0) / 7465\\n\\n\\nprint(f\\\"g_tot_acc: {g_tot_acc * 100} (%)\\\")\\nprint(f\\\"bb_tot_acc: {bb_tot_acc * 100} (%)\\\")\\nprint(f\\\"g_tot_f1: {g_tot_f1 * 100} (%)\\\")\\nprint(f\\\"g_tot_precision: {g_tot_precision * 100} (%)\\\")\\nprint(f\\\"g_tot_recall: {g_tot_recall * 100} (%)\\\")\\nprint(f\\\"bb_tot_recall: {bb_tot_recall * 100} (%)\\\")\\nprint(f\\\"total samples covered by g: {preds.size(0)} (out of {7465})\\\")\\nprint(f\\\"total coverage by g: {g_tot_cov}\\\")\";\n",
       "                var nbb_formatted_code = \"expert_type = \\\"explainer\\\"\\noutput = \\\"g_outputs\\\"\\n\\npreds = torch.cat(\\n    (\\n        test_tensor_preds_1,\\n        test_tensor_preds_2,\\n        test_tensor_preds_3,\\n        test_tensor_preds_4,\\n        test_tensor_preds_5,\\n        test_tensor_preds_6,\\n        test_tensor_preds_residual,\\n        #                    test_tensor_preds_7, test_tensor_preds_8\\n    ),\\n    dim=0,\\n)\\ngt = torch.cat(\\n    (\\n        test_tensor_y_1,\\n        test_tensor_y_2,\\n        test_tensor_y_3,\\n        test_tensor_y_4,\\n        test_tensor_y_5,\\n        test_tensor_y_6,\\n        test_tensor_y_residual\\n        #                test_tensor_y_7, test_tensor_y_8\\n    ),\\n    dim=0,\\n)\\nbb = torch.cat(\\n    (\\n        test_tensor_preds_bb_1.cpu(),\\n        test_tensor_preds_bb_2.cpu(),\\n        test_tensor_preds_bb_3.cpu(),\\n        test_tensor_preds_bb_4.cpu(),\\n        test_tensor_preds_bb_5.cpu(),\\n        test_tensor_preds_bb_6.cpu(),\\n        test_tensor_preds_bb_residual.cpu()\\n        #     test_tensor_preds_bb_7.cpu(), test_tensor_preds_bb_8.cpu()\\n    ),\\n    dim=0,\\n)\\n\\npreds_tot_np = preds.argmax(dim=1).numpy()\\ngt_tot_np = gt.numpy()\\nbb_tot_np = bb.argmax(dim=1).numpy()\\nbb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\\ng_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\\ng_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\ng_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\\\"micro\\\")\\nbb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\\\"micro\\\")\\ng_tot_cov = preds.size(0) / 7465\\n\\n\\nprint(f\\\"g_tot_acc: {g_tot_acc * 100} (%)\\\")\\nprint(f\\\"bb_tot_acc: {bb_tot_acc * 100} (%)\\\")\\nprint(f\\\"g_tot_f1: {g_tot_f1 * 100} (%)\\\")\\nprint(f\\\"g_tot_precision: {g_tot_precision * 100} (%)\\\")\\nprint(f\\\"g_tot_recall: {g_tot_recall * 100} (%)\\\")\\nprint(f\\\"bb_tot_recall: {bb_tot_recall * 100} (%)\\\")\\nprint(f\\\"total samples covered by g: {preds.size(0)} (out of {7465})\\\")\\nprint(f\\\"total coverage by g: {g_tot_cov}\\\")\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "expert_type = \"explainer\"\n",
    "output = \"g_outputs\"\n",
    "\n",
    "preds = torch.cat((test_tensor_preds_1, test_tensor_preds_2, test_tensor_preds_3, \n",
    "                   test_tensor_preds_4, test_tensor_preds_5, test_tensor_preds_6,\n",
    "                     test_tensor_preds_residual,\n",
    "#                    test_tensor_preds_7, test_tensor_preds_8\n",
    "                  ), dim=0)\n",
    "gt = torch.cat((test_tensor_y_1, test_tensor_y_2, test_tensor_y_3 ,\n",
    "                test_tensor_y_4, test_tensor_y_5, test_tensor_y_6,\n",
    "                test_tensor_y_residual\n",
    "#                test_tensor_y_7, test_tensor_y_8\n",
    "               ), dim=0)\n",
    "bb = torch.cat((\n",
    "    test_tensor_preds_bb_1.cpu(), test_tensor_preds_bb_2.cpu(), test_tensor_preds_bb_3.cpu(),\n",
    "    test_tensor_preds_bb_4.cpu(), test_tensor_preds_bb_5.cpu(), test_tensor_preds_bb_6.cpu(),\n",
    "    test_tensor_preds_bb_residual.cpu()\n",
    "#     test_tensor_preds_bb_7.cpu(), test_tensor_preds_bb_8.cpu()\n",
    "), \n",
    "               dim=0)\n",
    "\n",
    "preds_tot_np = preds.argmax(dim=1).numpy()\n",
    "gt_tot_np = gt.numpy()\n",
    "bb_tot_np = bb.argmax(dim=1).numpy()\n",
    "bb_tot_acc = metrics.accuracy_score(gt_tot_np, bb_tot_np)\n",
    "g_tot_acc = metrics.accuracy_score(gt_tot_np, preds_tot_np)\n",
    "g_tot_f1 = metrics.f1_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "g_tot_precision = metrics.precision_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "g_tot_recall = metrics.recall_score(gt_tot_np, preds_tot_np, average=\"micro\")\n",
    "bb_tot_recall = metrics.recall_score(gt_tot_np, bb_tot_np, average=\"micro\")\n",
    "g_tot_cov = preds.size(0) / 7465\n",
    "\n",
    "\n",
    "print(f\"g_tot_acc: {g_tot_acc * 100} (%)\")\n",
    "print(f\"bb_tot_acc: {bb_tot_acc * 100} (%)\")\n",
    "print(f\"g_tot_f1: {g_tot_f1 * 100} (%)\")\n",
    "print(f\"g_tot_precision: {g_tot_precision * 100} (%)\")\n",
    "print(f\"g_tot_recall: {g_tot_recall * 100} (%)\")\n",
    "print(f\"bb_tot_recall: {bb_tot_recall * 100} (%)\")\n",
    "print(f\"total samples covered by g: {preds.size(0)} (out of {7465})\")\n",
    "print(f\"total coverage by g: {g_tot_cov}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.18995311871839704,\n",
       " 0.21219022757880893,\n",
       " 0.18258540280617408,\n",
       " 0.1338245158856616,\n",
       " 0.1209644974595496,\n",
       " 0.08667112965753074,\n",
       " 0.04902880043290401)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 14;\n",
       "                var nbb_unformatted_code = \"(\\n    (\\n        torch.sum(test_tensor_preds_1.argmax(dim=1) == test_tensor_y_1)\\n        / test_tensor_y_1.size(0)\\n    ).item()\\n    * (test_tensor_preds_1.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_2.argmax(dim=1) == test_tensor_y_2)\\n        / test_tensor_y_2.size(0)\\n    ).item()\\n    * (test_tensor_preds_2.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_3.argmax(dim=1) == test_tensor_y_3)\\n        / test_tensor_y_3.size(0)\\n    ).item()\\n    * (test_tensor_preds_3.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_4.argmax(dim=1) == test_tensor_y_4)\\n        / test_tensor_y_4.size(0)\\n    ).item()\\n    * (test_tensor_preds_4.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_5.argmax(dim=1) == test_tensor_y_5)\\n        / test_tensor_y_5.size(0)\\n    ).item()\\n    * (test_tensor_preds_5.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_6.argmax(dim=1) == test_tensor_y_6)\\n        / test_tensor_y_6.size(0)\\n    ).item()\\n    * (test_tensor_preds_6.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual)\\n        / test_tensor_preds_residual.size(0)\\n    ).item()\\n    * (test_tensor_preds_residual.size(0) / 7465),\\n)\";\n",
       "                var nbb_formatted_code = \"(\\n    (\\n        torch.sum(test_tensor_preds_1.argmax(dim=1) == test_tensor_y_1)\\n        / test_tensor_y_1.size(0)\\n    ).item()\\n    * (test_tensor_preds_1.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_2.argmax(dim=1) == test_tensor_y_2)\\n        / test_tensor_y_2.size(0)\\n    ).item()\\n    * (test_tensor_preds_2.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_3.argmax(dim=1) == test_tensor_y_3)\\n        / test_tensor_y_3.size(0)\\n    ).item()\\n    * (test_tensor_preds_3.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_4.argmax(dim=1) == test_tensor_y_4)\\n        / test_tensor_y_4.size(0)\\n    ).item()\\n    * (test_tensor_preds_4.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_5.argmax(dim=1) == test_tensor_y_5)\\n        / test_tensor_y_5.size(0)\\n    ).item()\\n    * (test_tensor_preds_5.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_6.argmax(dim=1) == test_tensor_y_6)\\n        / test_tensor_y_6.size(0)\\n    ).item()\\n    * (test_tensor_preds_6.size(0) / 7465),\\n    (\\n        torch.sum(test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual)\\n        / test_tensor_preds_residual.size(0)\\n    ).item()\\n    * (test_tensor_preds_residual.size(0) / 7465),\\n)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_1.argmax(dim=1) == test_tensor_y_1)\n",
    "        / test_tensor_y_1.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_1.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_2.argmax(dim=1) == test_tensor_y_2)\n",
    "        / test_tensor_y_2.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_2.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_3.argmax(dim=1) == test_tensor_y_3)\n",
    "        / test_tensor_y_3.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_3.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_4.argmax(dim=1) == test_tensor_y_4)\n",
    "        / test_tensor_y_4.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_4.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_5.argmax(dim=1) == test_tensor_y_5)\n",
    "        / test_tensor_y_5.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_5.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_6.argmax(dim=1) == test_tensor_y_6)\n",
    "        / test_tensor_y_6.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_6.size(0) / 7465),\n",
    "    (\n",
    "        torch.sum(test_tensor_preds_residual.argmax(dim=1) == test_tensor_y_residual)\n",
    "        / test_tensor_preds_residual.size(0)\n",
    "    ).item()\n",
    "    * (test_tensor_preds_residual.size(0) / 7465),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.975217692539026"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 15;\n",
       "                var nbb_unformatted_code = \"(\\n    0.18995311871839704\\n    + 0.21219022757880893\\n    + 0.18258540280617408\\n    + 0.1338245158856616\\n    + 0.1209644974595496\\n    + 0.08667112965753074\\n    + 0.04902880043290401\\n)\";\n",
       "                var nbb_formatted_code = \"(\\n    0.18995311871839704\\n    + 0.21219022757880893\\n    + 0.18258540280617408\\n    + 0.1338245158856616\\n    + 0.1209644974595496\\n    + 0.08667112965753074\\n    + 0.04902880043290401\\n)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(\n",
    "    0.18995311871839704\n",
    "    + 0.21219022757880893\n",
    "    + 0.18258540280617408\n",
    "    + 0.1338245158856616\n",
    "    + 0.1209644974595496\n",
    "    + 0.08667112965753074\n",
    "    + 0.04902880043290401\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(50, 6)\n",
      "(50, 6)\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 16;\n",
       "                var nbb_unformatted_code = \"expert_1_gt, expert_1_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_1, pred_tensor=test_tensor_preds_1, n_classes=50\\n)\\n\\nexpert_2_gt, expert_2_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_2, pred_tensor=test_tensor_preds_2, n_classes=50\\n)\\nexpert_3_gt, expert_3_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_3, pred_tensor=test_tensor_preds_3, n_classes=50\\n)\\nexpert_4_gt, expert_4_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_4, pred_tensor=test_tensor_preds_4, n_classes=50\\n)\\nexpert_5_gt, expert_5_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_5, pred_tensor=test_tensor_preds_5, n_classes=50\\n)\\nexpert_6_gt, expert_6_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_6, pred_tensor=test_tensor_preds_6, n_classes=50\\n)\\n# expert_7_gt, expert_7_correct_pred = plot.create_data_for_expert_label_plot(\\n#     GT_tensor=test_tensor_y_7, pred_tensor=test_tensor_preds_7, n_classes=200\\n# )\\n# expert_8_gt, expert_8_correct_pred = plot.create_data_for_expert_label_plot(\\n#     GT_tensor=test_tensor_y_8, pred_tensor=test_tensor_preds_8, n_classes=200\\n# )\\n\\n\\nexpert_1_gt =np.array([expert_1_gt]).T\\nexpert_2_gt =np.array([expert_2_gt]).T\\nexpert_3_gt =np.array([expert_3_gt]).T\\nexpert_4_gt =np.array([expert_4_gt]).T\\nexpert_5_gt =np.array([expert_5_gt]).T\\nexpert_6_gt =np.array([expert_6_gt]).T\\n# expert_7_gt =np.array([expert_7_gt]).T\\n# expert_8_gt =np.array([expert_8_gt]).T\\n\\nexpert_gt = np.concatenate(\\n    (\\n        expert_1_gt, \\n        expert_2_gt,\\n        expert_3_gt,\\n        expert_4_gt,\\n        expert_5_gt,\\n        expert_6_gt,\\n#         expert_7_gt,\\n#         expert_8_gt\\n    ), axis=1)\\n\\nexpert_1_correct_pred =np.array([expert_1_correct_pred]).T\\nexpert_2_correct_pred =np.array([expert_2_correct_pred]).T\\nexpert_3_correct_pred =np.array([expert_3_correct_pred]).T\\nexpert_4_correct_pred =np.array([expert_4_correct_pred]).T\\nexpert_5_correct_pred =np.array([expert_5_correct_pred]).T\\nexpert_6_correct_pred =np.array([expert_6_correct_pred]).T\\n# expert_7_correct_pred =np.array([expert_7_correct_pred]).T\\n# expert_8_correct_pred =np.array([expert_8_correct_pred]).T\\n\\nexpert_correct_preds = np.concatenate(\\n    (\\n        expert_1_correct_pred, \\n        expert_2_correct_pred,\\n        expert_3_correct_pred,\\n        expert_4_correct_pred,\\n        expert_5_correct_pred,\\n        expert_6_correct_pred,\\n#         expert_7_correct_pred,\\n#         expert_8_correct_pred\\n    ), axis=1)\\nprint(expert_gt.shape)\\nprint(expert_correct_preds.shape)\\n\\n\\n\\n\\n\\n\\n# print(exper_gt.shape)\\n# print(expert_correct_preds.shape)\";\n",
       "                var nbb_formatted_code = \"expert_1_gt, expert_1_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_1, pred_tensor=test_tensor_preds_1, n_classes=50\\n)\\n\\nexpert_2_gt, expert_2_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_2, pred_tensor=test_tensor_preds_2, n_classes=50\\n)\\nexpert_3_gt, expert_3_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_3, pred_tensor=test_tensor_preds_3, n_classes=50\\n)\\nexpert_4_gt, expert_4_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_4, pred_tensor=test_tensor_preds_4, n_classes=50\\n)\\nexpert_5_gt, expert_5_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_5, pred_tensor=test_tensor_preds_5, n_classes=50\\n)\\nexpert_6_gt, expert_6_correct_pred = plot.create_data_for_expert_label_plot(\\n    GT_tensor=test_tensor_y_6, pred_tensor=test_tensor_preds_6, n_classes=50\\n)\\n# expert_7_gt, expert_7_correct_pred = plot.create_data_for_expert_label_plot(\\n#     GT_tensor=test_tensor_y_7, pred_tensor=test_tensor_preds_7, n_classes=200\\n# )\\n# expert_8_gt, expert_8_correct_pred = plot.create_data_for_expert_label_plot(\\n#     GT_tensor=test_tensor_y_8, pred_tensor=test_tensor_preds_8, n_classes=200\\n# )\\n\\n\\nexpert_1_gt = np.array([expert_1_gt]).T\\nexpert_2_gt = np.array([expert_2_gt]).T\\nexpert_3_gt = np.array([expert_3_gt]).T\\nexpert_4_gt = np.array([expert_4_gt]).T\\nexpert_5_gt = np.array([expert_5_gt]).T\\nexpert_6_gt = np.array([expert_6_gt]).T\\n# expert_7_gt =np.array([expert_7_gt]).T\\n# expert_8_gt =np.array([expert_8_gt]).T\\n\\nexpert_gt = np.concatenate(\\n    (\\n        expert_1_gt,\\n        expert_2_gt,\\n        expert_3_gt,\\n        expert_4_gt,\\n        expert_5_gt,\\n        expert_6_gt,\\n        #         expert_7_gt,\\n        #         expert_8_gt\\n    ),\\n    axis=1,\\n)\\n\\nexpert_1_correct_pred = np.array([expert_1_correct_pred]).T\\nexpert_2_correct_pred = np.array([expert_2_correct_pred]).T\\nexpert_3_correct_pred = np.array([expert_3_correct_pred]).T\\nexpert_4_correct_pred = np.array([expert_4_correct_pred]).T\\nexpert_5_correct_pred = np.array([expert_5_correct_pred]).T\\nexpert_6_correct_pred = np.array([expert_6_correct_pred]).T\\n# expert_7_correct_pred =np.array([expert_7_correct_pred]).T\\n# expert_8_correct_pred =np.array([expert_8_correct_pred]).T\\n\\nexpert_correct_preds = np.concatenate(\\n    (\\n        expert_1_correct_pred,\\n        expert_2_correct_pred,\\n        expert_3_correct_pred,\\n        expert_4_correct_pred,\\n        expert_5_correct_pred,\\n        expert_6_correct_pred,\\n        #         expert_7_correct_pred,\\n        #         expert_8_correct_pred\\n    ),\\n    axis=1,\\n)\\nprint(expert_gt.shape)\\nprint(expert_correct_preds.shape)\\n\\n\\n# print(exper_gt.shape)\\n# print(expert_correct_preds.shape)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "expert_1_gt, expert_1_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_1, pred_tensor=test_tensor_preds_1, n_classes=50\n",
    ")\n",
    "\n",
    "expert_2_gt, expert_2_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_2, pred_tensor=test_tensor_preds_2, n_classes=50\n",
    ")\n",
    "expert_3_gt, expert_3_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_3, pred_tensor=test_tensor_preds_3, n_classes=50\n",
    ")\n",
    "expert_4_gt, expert_4_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_4, pred_tensor=test_tensor_preds_4, n_classes=50\n",
    ")\n",
    "expert_5_gt, expert_5_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_5, pred_tensor=test_tensor_preds_5, n_classes=50\n",
    ")\n",
    "expert_6_gt, expert_6_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "    GT_tensor=test_tensor_y_6, pred_tensor=test_tensor_preds_6, n_classes=50\n",
    ")\n",
    "# expert_7_gt, expert_7_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "#     GT_tensor=test_tensor_y_7, pred_tensor=test_tensor_preds_7, n_classes=200\n",
    "# )\n",
    "# expert_8_gt, expert_8_correct_pred = plot.create_data_for_expert_label_plot(\n",
    "#     GT_tensor=test_tensor_y_8, pred_tensor=test_tensor_preds_8, n_classes=200\n",
    "# )\n",
    "\n",
    "\n",
    "expert_1_gt =np.array([expert_1_gt]).T\n",
    "expert_2_gt =np.array([expert_2_gt]).T\n",
    "expert_3_gt =np.array([expert_3_gt]).T\n",
    "expert_4_gt =np.array([expert_4_gt]).T\n",
    "expert_5_gt =np.array([expert_5_gt]).T\n",
    "expert_6_gt =np.array([expert_6_gt]).T\n",
    "# expert_7_gt =np.array([expert_7_gt]).T\n",
    "# expert_8_gt =np.array([expert_8_gt]).T\n",
    "\n",
    "expert_gt = np.concatenate(\n",
    "    (\n",
    "        expert_1_gt, \n",
    "        expert_2_gt,\n",
    "        expert_3_gt,\n",
    "        expert_4_gt,\n",
    "        expert_5_gt,\n",
    "        expert_6_gt,\n",
    "#         expert_7_gt,\n",
    "#         expert_8_gt\n",
    "    ), axis=1)\n",
    "\n",
    "expert_1_correct_pred =np.array([expert_1_correct_pred]).T\n",
    "expert_2_correct_pred =np.array([expert_2_correct_pred]).T\n",
    "expert_3_correct_pred =np.array([expert_3_correct_pred]).T\n",
    "expert_4_correct_pred =np.array([expert_4_correct_pred]).T\n",
    "expert_5_correct_pred =np.array([expert_5_correct_pred]).T\n",
    "expert_6_correct_pred =np.array([expert_6_correct_pred]).T\n",
    "# expert_7_correct_pred =np.array([expert_7_correct_pred]).T\n",
    "# expert_8_correct_pred =np.array([expert_8_correct_pred]).T\n",
    "\n",
    "expert_correct_preds = np.concatenate(\n",
    "    (\n",
    "        expert_1_correct_pred, \n",
    "        expert_2_correct_pred,\n",
    "        expert_3_correct_pred,\n",
    "        expert_4_correct_pred,\n",
    "        expert_5_correct_pred,\n",
    "        expert_6_correct_pred,\n",
    "#         expert_7_correct_pred,\n",
    "#         expert_8_correct_pred\n",
    "    ), axis=1)\n",
    "print(expert_gt.shape)\n",
    "print(expert_correct_preds.shape)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# print(exper_gt.shape)\n",
    "# print(expert_correct_preds.shape)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[13 25 19 26 46 49  2  8  9 17  5 14 41 23  0 39 27 47 20  1 44 36 30  3\n",
      " 10 12 28 37 24 15 32  6  4 38 45  7 18 16 40 21 35 31 48 29 11 43 34 33\n",
      " 42 22]\n",
      "50\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[  0,   0,   0,   0,   0,   0],\n",
       "       [  1,   0,   0,   0,   0,   0],\n",
       "       [174,   0,   0,   0,   0,   0],\n",
       "       [240,   0,   0,   0,   0,   0],\n",
       "       [ 42,   1,   0,   0,   0,   0],\n",
       "       [188,   1,   0,   0,   0,   0],\n",
       "       [ 56,   2,   0,   0,   0,   0],\n",
       "       [  2,  33,   0,   0,   0,   0],\n",
       "       [  0,  92,   0,   0,   0,   0],\n",
       "       [  6, 136,   0,   0,   0,   0],\n",
       "       [  0, 143,   0,   0,   0,   0],\n",
       "       [  0, 144,   0,   0,   0,   0],\n",
       "       [  0, 136,   1,   0,   0,   0],\n",
       "       [ 43, 154,   1,   0,   0,   0],\n",
       "       [  0, 208,   1,   0,   0,   0],\n",
       "       [  0, 268,   1,   0,   0,   0],\n",
       "       [135,   0,   4,   0,   0,   0],\n",
       "       [ 92,   0,   6,   0,   0,   0],\n",
       "       [  1,   0, 145,   0,   0,   0],\n",
       "       [  0,   0, 170,   0,   0,   0],\n",
       "       [  0,   1, 173,   0,   0,   0],\n",
       "       [  0,   0, 179,   0,   0,   0],\n",
       "       [  0,   0, 240,   0,   0,   0],\n",
       "       [ 23,   0,   2,  12,   0,   0],\n",
       "       [  0,   0,   0,  37,   0,   0],\n",
       "       [  0,   0,   0, 175,   0,   0],\n",
       "       [  0,   0,   0, 217,   0,   0],\n",
       "       [  0,   0,   1, 233,   0,   0],\n",
       "       [144,   0,   1,   0,   1,   0],\n",
       "       [  0,  40,  98,   1,   1,   0],\n",
       "       [ 67,   0,   0,   0,   2,   0],\n",
       "       [  0,   0,   2, 325,   2,   0],\n",
       "       [  0, 107,   0,   0,   3,   0],\n",
       "       [  0,   0,   0,   0, 175,   0],\n",
       "       [  0,   0,   0,   0, 204,   0],\n",
       "       [  0,   0,   0,   0, 206,   0],\n",
       "       [  0,   0,   0,   0, 207,   0],\n",
       "       [ 52,   5,   0,   0,   0,   1],\n",
       "       [  0, 122,   3,   0,   0,   1],\n",
       "       [  0,   0, 132,   0,   0,   1],\n",
       "       [ 98,  50,   2,   1,   0,   1],\n",
       "       [  0,   0,   1,   0, 116,   1],\n",
       "       [  0,   0, 265,   0,   1,   2],\n",
       "       [ 72,   0,   0,   0,   0,   4],\n",
       "       [  0,   0,   0,   0,   0,  20],\n",
       "       [  0,   0,   0,   0,   0,  28],\n",
       "       [  0,   0,   2,   1,   0,  51],\n",
       "       [  0,   0,   0,   0,   0,  59],\n",
       "       [  0,   0,   0,   0,   0, 204],\n",
       "       [  0,   0,   2,   0,   0, 282]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 17;\n",
       "                var nbb_unformatted_code = \"order = np.lexsort(\\n    [\\n        expert_gt[:, 0],\\n        expert_gt[:, 1],\\n        expert_gt[:, 2],\\n        expert_gt[:, 3],\\n        expert_gt[:, 4],\\n        expert_gt[:, 5],\\n#         expert_gt[:, 6],\\n#         expert_gt[:, 7],\\n    ]\\n)\\nprint(order)\\nprint(len(order))\\nsorted_cols = []\\nfor idx in order:\\n    sorted_cols.append(args.labels[idx])\\n# print(args.labels)\\n# print(sorted_cols)\\nexpert_gt[order]\";\n",
       "                var nbb_formatted_code = \"order = np.lexsort(\\n    [\\n        expert_gt[:, 0],\\n        expert_gt[:, 1],\\n        expert_gt[:, 2],\\n        expert_gt[:, 3],\\n        expert_gt[:, 4],\\n        expert_gt[:, 5],\\n        #         expert_gt[:, 6],\\n        #         expert_gt[:, 7],\\n    ]\\n)\\nprint(order)\\nprint(len(order))\\nsorted_cols = []\\nfor idx in order:\\n    sorted_cols.append(args.labels[idx])\\n# print(args.labels)\\n# print(sorted_cols)\\nexpert_gt[order]\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "order = np.lexsort(\n",
    "    [\n",
    "        expert_gt[:, 0],\n",
    "        expert_gt[:, 1],\n",
    "        expert_gt[:, 2],\n",
    "        expert_gt[:, 3],\n",
    "        expert_gt[:, 4],\n",
    "        expert_gt[:, 5],\n",
    "#         expert_gt[:, 6],\n",
    "#         expert_gt[:, 7],\n",
    "    ]\n",
    ")\n",
    "print(order)\n",
    "print(len(order))\n",
    "sorted_cols = []\n",
    "for idx in order:\n",
    "    sorted_cols.append(args.labels[idx])\n",
    "# print(args.labels)\n",
    "# print(sorted_cols)\n",
    "expert_gt[order]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'y_axis_labels' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_102553/2088249758.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;31m# print(y_axis_labels)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0;32mfor\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_axis_labels\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m     \u001b[0my_axis_labels\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0midx\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreplace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"+\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\" \"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0;31m#     print(idx, label)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'y_axis_labels' is not defined"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 18;\n",
       "                var nbb_unformatted_code = \"import matplotlib.pyplot as plt\\nimport seaborn as sns\\n\\n# print(y_axis_labels)\\nfor idx, label in enumerate(y_axis_labels):\\n    y_axis_labels[idx] = label.replace(\\\"+\\\", \\\" \\\")\\n#     print(idx, label)\\n\\n\\ny_axis_labels = sorted_cols\\nx_axis_labels = [1, 2, 3, 4, 5, 6]\\n# for i in range(1, 201):\\n#     y_axis_labels.append(i)\\n\\nfig, ax = plt.subplots(figsize=(12, 12))\\nsns.set(font_scale=1.2)\\np = sns.heatmap(\\n    expert_gt[order],\\n    xticklabels=x_axis_labels,\\n    yticklabels=y_axis_labels,\\n    linewidths=0.1,\\n    cmap=\\\"BuPu\\\",\\n)\\np.set_xlabel(\\\"# Experts\\\")\\np.set_ylabel(\\\"Class Labels\\\")\\n# p.set_yticklabels(p.get_yticklabels(), rotation=10)\\nplt.show()\";\n",
       "                var nbb_formatted_code = \"import matplotlib.pyplot as plt\\nimport seaborn as sns\\n\\n# print(y_axis_labels)\\nfor idx, label in enumerate(y_axis_labels):\\n    y_axis_labels[idx] = label.replace(\\\"+\\\", \\\" \\\")\\n#     print(idx, label)\\n\\n\\ny_axis_labels = sorted_cols\\nx_axis_labels = [1, 2, 3, 4, 5, 6]\\n# for i in range(1, 201):\\n#     y_axis_labels.append(i)\\n\\nfig, ax = plt.subplots(figsize=(12, 12))\\nsns.set(font_scale=1.2)\\np = sns.heatmap(\\n    expert_gt[order],\\n    xticklabels=x_axis_labels,\\n    yticklabels=y_axis_labels,\\n    linewidths=0.1,\\n    cmap=\\\"BuPu\\\",\\n)\\np.set_xlabel(\\\"# Experts\\\")\\np.set_ylabel(\\\"Class Labels\\\")\\n# p.set_yticklabels(p.get_yticklabels(), rotation=10)\\nplt.show()\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# print(y_axis_labels)\n",
    "for idx, label in enumerate(y_axis_labels):\n",
    "    y_axis_labels[idx] = label.replace(\"+\", \" \")\n",
    "#     print(idx, label)\n",
    "\n",
    "\n",
    "y_axis_labels = sorted_cols\n",
    "x_axis_labels = [1, 2, 3, 4, 5, 6]\n",
    "# for i in range(1, 201):\n",
    "#     y_axis_labels.append(i)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 12))\n",
    "sns.set(font_scale=1.2)\n",
    "p = sns.heatmap(\n",
    "    expert_gt[order],\n",
    "    xticklabels=x_axis_labels,\n",
    "    yticklabels=y_axis_labels,\n",
    "    linewidths=0.1,\n",
    "    cmap=\"BuPu\",\n",
    ")\n",
    "p.set_xlabel(\"# Experts\")\n",
    "p.set_ylabel(\"Class Labels\")\n",
    "# p.set_yticklabels(p.get_yticklabels(), rotation=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[11 13 22 25 33 34 42 43 46 32 29 47 27 24 19 26  2 49  8 16 35  9  4 40\n",
      " 41 17  5 14 23 39  0 15 20 21  1 44 36 30 48  3 10 12 28 37  6 31 38 45\n",
      "  7 18]\n",
      "50\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 19;\n",
       "                var nbb_unformatted_code = \"import matplotlib.pyplot as plt\\nimport seaborn as sns\\n\\norder = np.lexsort(\\n    [\\n        expert_correct_preds[:, 0],\\n        expert_correct_preds[:, 1],\\n        expert_correct_preds[:, 2],\\n        expert_correct_preds[:, 3],\\n        expert_correct_preds[:, 4],\\n        #         expert_correct_preds[:, 5],\\n        #         expert_correct_preds[:, 6],\\n        #         expert_correct_preds[:, 7],\\n    ],\\n    axis=0,\\n)\\n\\n\\n# print(y_axis_labels)\\nprint(order)\\nprint(len(order))\\nsorted_cols = []\\nfor idx in order:\\n    sorted_cols.append(args.labels[idx])\\n# print(args.labels)\\n# print(sorted_cols)\\n\\nplot_folder = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/Plots/Appendix\\\"\\n\\ny_axis_labels = sorted_cols\\nfor idx, label in enumerate(y_axis_labels):\\n    y_axis_labels[idx] = label.replace(\\\"+\\\", \\\" \\\")\\n# x_axis_labels = [1, 2, 3, 4, 5, ]\\nx_axis_labels = [\\n    1,\\n    2,\\n    3,\\n    4,\\n    5,\\n]\\n# for i in range(1, 201):\\n#     y_axis_labels.append(i)\\n\\nplt.figure().clear()\\nplt.close()\\nplt.cla()\\nplt.clf()\\nsns.reset_orig()\\nfig, ax = plt.subplots(figsize=(15, 15))\\n# sns.set(font_scale=3)\\n# p = sns.heatmap(\\n#     expert_correct_preds[order],\\n#     xticklabels=x_axis_labels,\\n#     yticklabels=y_axis_labels,\\n#     linewidths=0.5,\\n# )\\n\\np = sns.heatmap(\\n    expert_gt[order],\\n    xticklabels=x_axis_labels,\\n    yticklabels=y_axis_labels,\\n    linewidths=0.1,\\n    cmap=\\\"BuPu\\\",\\n)\\nax.set_xticklabels(x_axis_labels, fontsize=20)\\nax.set_yticklabels(y_axis_labels, fontsize=20)\\np.set_xlabel(\\\"# Experts\\\", fontsize=35)\\np.set_ylabel(\\\"Class Labels\\\", fontsize=35)\\n\\nresolution_value = 500\\n# plt.savefig(\\n#     os.path.join(plot_folder, \\\"VIT-CUB-performance.png\\\"),\\n#     format=\\\"png\\\",\\n#     dpi=resolution_value,\\n#     bbox_inches=\\\"tight\\\",\\n# )\\n# p.set_yticklabels(p.get_yticklabels(), rotation=10)\\nplt.show()\";\n",
       "                var nbb_formatted_code = \"import matplotlib.pyplot as plt\\nimport seaborn as sns\\n\\norder = np.lexsort(\\n    [\\n        expert_correct_preds[:, 0],\\n        expert_correct_preds[:, 1],\\n        expert_correct_preds[:, 2],\\n        expert_correct_preds[:, 3],\\n        expert_correct_preds[:, 4],\\n        #         expert_correct_preds[:, 5],\\n        #         expert_correct_preds[:, 6],\\n        #         expert_correct_preds[:, 7],\\n    ],\\n    axis=0,\\n)\\n\\n\\n# print(y_axis_labels)\\nprint(order)\\nprint(len(order))\\nsorted_cols = []\\nfor idx in order:\\n    sorted_cols.append(args.labels[idx])\\n# print(args.labels)\\n# print(sorted_cols)\\n\\nplot_folder = \\\"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/Plots/Appendix\\\"\\n\\ny_axis_labels = sorted_cols\\nfor idx, label in enumerate(y_axis_labels):\\n    y_axis_labels[idx] = label.replace(\\\"+\\\", \\\" \\\")\\n# x_axis_labels = [1, 2, 3, 4, 5, ]\\nx_axis_labels = [\\n    1,\\n    2,\\n    3,\\n    4,\\n    5,\\n]\\n# for i in range(1, 201):\\n#     y_axis_labels.append(i)\\n\\nplt.figure().clear()\\nplt.close()\\nplt.cla()\\nplt.clf()\\nsns.reset_orig()\\nfig, ax = plt.subplots(figsize=(15, 15))\\n# sns.set(font_scale=3)\\n# p = sns.heatmap(\\n#     expert_correct_preds[order],\\n#     xticklabels=x_axis_labels,\\n#     yticklabels=y_axis_labels,\\n#     linewidths=0.5,\\n# )\\n\\np = sns.heatmap(\\n    expert_gt[order],\\n    xticklabels=x_axis_labels,\\n    yticklabels=y_axis_labels,\\n    linewidths=0.1,\\n    cmap=\\\"BuPu\\\",\\n)\\nax.set_xticklabels(x_axis_labels, fontsize=20)\\nax.set_yticklabels(y_axis_labels, fontsize=20)\\np.set_xlabel(\\\"# Experts\\\", fontsize=35)\\np.set_ylabel(\\\"Class Labels\\\", fontsize=35)\\n\\nresolution_value = 500\\n# plt.savefig(\\n#     os.path.join(plot_folder, \\\"VIT-CUB-performance.png\\\"),\\n#     format=\\\"png\\\",\\n#     dpi=resolution_value,\\n#     bbox_inches=\\\"tight\\\",\\n# )\\n# p.set_yticklabels(p.get_yticklabels(), rotation=10)\\nplt.show()\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "order = np.lexsort(\n",
    "    [\n",
    "        expert_correct_preds[:, 0],\n",
    "        expert_correct_preds[:, 1],\n",
    "        expert_correct_preds[:, 2],\n",
    "        expert_correct_preds[:, 3],\n",
    "        expert_correct_preds[:, 4],\n",
    "        #         expert_correct_preds[:, 5],\n",
    "        #         expert_correct_preds[:, 6],\n",
    "        #         expert_correct_preds[:, 7],\n",
    "    ],\n",
    "    axis=0,\n",
    ")\n",
    "\n",
    "\n",
    "# print(y_axis_labels)\n",
    "print(order)\n",
    "print(len(order))\n",
    "sorted_cols = []\n",
    "for idx in order:\n",
    "    sorted_cols.append(args.labels[idx])\n",
    "# print(args.labels)\n",
    "# print(sorted_cols)\n",
    "\n",
    "plot_folder = \"/ocean/projects/asc170022p/shg121/PhD/ICLR-2022/Plots/Appendix\"\n",
    "\n",
    "y_axis_labels = sorted_cols\n",
    "for idx, label in enumerate(y_axis_labels):\n",
    "    y_axis_labels[idx] = label.replace(\"+\", \" \")\n",
    "# x_axis_labels = [1, 2, 3, 4, 5, ]\n",
    "x_axis_labels = [\n",
    "    1,\n",
    "    2,\n",
    "    3,\n",
    "    4,\n",
    "    5,\n",
    "]\n",
    "# for i in range(1, 201):\n",
    "#     y_axis_labels.append(i)\n",
    "\n",
    "plt.figure().clear()\n",
    "plt.close()\n",
    "plt.cla()\n",
    "plt.clf()\n",
    "sns.reset_orig()\n",
    "fig, ax = plt.subplots(figsize=(15, 15))\n",
    "# sns.set(font_scale=3)\n",
    "# p = sns.heatmap(\n",
    "#     expert_correct_preds[order],\n",
    "#     xticklabels=x_axis_labels,\n",
    "#     yticklabels=y_axis_labels,\n",
    "#     linewidths=0.5,\n",
    "# )\n",
    "\n",
    "p = sns.heatmap(\n",
    "    expert_gt[order],\n",
    "    xticklabels=x_axis_labels,\n",
    "    yticklabels=y_axis_labels,\n",
    "    linewidths=0.1,\n",
    "    cmap=\"BuPu\",\n",
    ")\n",
    "ax.set_xticklabels(x_axis_labels, fontsize=20)\n",
    "ax.set_yticklabels(y_axis_labels, fontsize=20)\n",
    "p.set_xlabel(\"# Experts\", fontsize=35)\n",
    "p.set_ylabel(\"Class Labels\", fontsize=35)\n",
    "\n",
    "resolution_value = 500\n",
    "# plt.savefig(\n",
    "#     os.path.join(plot_folder, \"VIT-CUB-performance.png\"),\n",
    "#     format=\"png\",\n",
    "#     dpi=resolution_value,\n",
    "#     bbox_inches=\"tight\",\n",
    "# )\n",
    "# p.set_yticklabels(p.get_yticklabels(), rotation=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(0)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(0)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(0)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(0)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(15)\n",
      "tensor(21)\n",
      "tensor(15)\n",
      "tensor(30)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(1)\n",
      "tensor(44)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(30)\n",
      "tensor(1)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(36)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(15)\n",
      "tensor(15)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(21)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(15)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(21)\n",
      "tensor(1)\n",
      "tensor(20)\n",
      "tensor(30)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(30)\n",
      "tensor(30)\n",
      "tensor(44)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(1)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(44)\n",
      "tensor(48)\n",
      "tensor(1)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(21)\n",
      "tensor(20)\n",
      "tensor(15)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(20)\n",
      "tensor(20)\n",
      "tensor(44)\n",
      "tensor(20)\n",
      "tensor(36)\n",
      "tensor(48)\n",
      "tensor(48)\n",
      "tensor(36)\n",
      "tensor(20)\n",
      "tensor(48)\n",
      "tensor(48)\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 47;\n",
       "                var nbb_unformatted_code = \"for i in test_tensor_preds_3.argmax(dim=1):\\n    print(i)\";\n",
       "                var nbb_formatted_code = \"for i in test_tensor_preds_3.argmax(dim=1):\\n    print(i)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in test_tensor_preds_3.argmax(dim=1):\n",
    "    print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(27.) rhinoceros\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(24.) chimpanzee\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(47.) raccoon\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(35.) otter\n",
      "tensor(47.) raccoon\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(47.) raccoon\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(22.) sheep\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(3.) beaver\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(34.) weasel\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(47.) raccoon\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(22.) sheep\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(40.) bobcat\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(41.) pig\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(23.) seal\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(35.) otter\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(40.) bobcat\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(37.) zebra\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(47.) raccoon\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(6.) horse\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(31.) wolf\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(6.) horse\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(39.) deer\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(27.) rhinoceros\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(3.) beaver\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(27.) rhinoceros\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(0.) antelope\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(34.) weasel\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(15.) moose\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(47.) raccoon\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(30.) giraffe\n",
      "tensor(21.) fox\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(15.) moose\n",
      "tensor(40.) bobcat\n",
      "tensor(48.) cow\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(30.) giraffe\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(36.) buffalo\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(15.) moose\n",
      "tensor(15.) moose\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(21.) fox\n",
      "tensor(21.) fox\n",
      "tensor(27.) rhinoceros\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(15.) moose\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(21.) fox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(20.) ox\n",
      "tensor(30.) giraffe\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(30.) giraffe\n",
      "tensor(30.) giraffe\n",
      "tensor(44.) polar+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(1.) grizzly+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(21.) fox\n",
      "tensor(48.) cow\n",
      "tensor(15.) moose\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(20.) ox\n",
      "tensor(20.) ox\n",
      "tensor(44.) polar+bear\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n",
      "tensor(36.) buffalo\n",
      "tensor(20.) ox\n",
      "tensor(48.) cow\n",
      "tensor(48.) cow\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 44;\n",
       "                var nbb_unformatted_code = \"for i in test_tensor_y_3:\\n    print(i, args.labels[int(i.item())])\";\n",
       "                var nbb_formatted_code = \"for i in test_tensor_y_3:\\n    print(i, args.labels[int(i.item())])\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in test_tensor_y_3:\n",
    "    print(i, args.labels[int(i.item())])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "python_3_7",
   "language": "python",
   "name": "python_3_7"
  },
  "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}