filter_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d2e3ce5f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/envs/bollworm/lib/python3.10/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "import gc\n",
    "import argparse\n",
    "import os\n",
    "import time\n",
    "import copy\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from collections import defaultdict\n",
    "\n",
    "# Sklearn\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "\n",
    "# PyTorch\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "\n",
    "os.environ[\"CUDA_LAUNCH_BLOCKING\"] = \"1\"\n",
    "\n",
    "# WandB\n",
    "import wandb\n",
    "\n",
    "# Custom Import\n",
    "from utils.loader import get_loaders\n",
    "from utils.train_utils import test\n",
    "from configs import filter_config\n",
    "from utils.models import build_model, load_model\n",
    "from utils.metric import get_metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9af81381",
   "metadata": {},
   "outputs": [],
   "source": [
    "CFG = filter_config.get_config()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9176a707",
   "metadata": {},
   "outputs": [],
   "source": [
    " # Create Folds\n",
    "train_filter = pd.read_csv(\"data/preproc/train_filter.csv\")\n",
    "fold =0\n",
    "val_df = train_filter.query(\"fold==@fold\").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "16e1b60e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_id_worm</th>\n",
       "      <th>bw</th>\n",
       "      <th>fold</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id_0005ef295aafe6acc63587db.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>id_00750938fb12fad4fd8fad24.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>id_009d7e2ef3c4f47b14c6c4c3.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>id_00e5892f475750f80a9f9ce5.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id_0128f4371f1b1a678062cff6.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1823</th>\n",
       "      <td>id_ff239c464c474c0936fd5d9c.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1824</th>\n",
       "      <td>id_ff52ef03790a72f1ca7c6837.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1825</th>\n",
       "      <td>id_ff6c6dc955223c93456ae3b9.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1826</th>\n",
       "      <td>id_ffc5b54765267f4aa7f015ca.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1827</th>\n",
       "      <td>id_fffdca893564aa86ff5be4cc.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1828 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        image_id_worm  bw  fold\n",
       "0     id_0005ef295aafe6acc63587db.jpg   1   0.0\n",
       "1     id_00750938fb12fad4fd8fad24.jpg   1   0.0\n",
       "2     id_009d7e2ef3c4f47b14c6c4c3.jpg   0   0.0\n",
       "3     id_00e5892f475750f80a9f9ce5.jpg   0   0.0\n",
       "4     id_0128f4371f1b1a678062cff6.jpg   1   0.0\n",
       "...                               ...  ..   ...\n",
       "1823  id_ff239c464c474c0936fd5d9c.jpg   0   0.0\n",
       "1824  id_ff52ef03790a72f1ca7c6837.jpg   1   0.0\n",
       "1825  id_ff6c6dc955223c93456ae3b9.jpg   1   0.0\n",
       "1826  id_ffc5b54765267f4aa7f015ca.jpg   1   0.0\n",
       "1827  id_fffdca893564aa86ff5be4cc.jpg   1   0.0\n",
       "\n",
       "[1828 rows x 3 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df =  val_df#pd.read_csv(\"data/preproc/val_df.csv\")\n",
    "test_df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "bff7bd6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_loader= get_loaders(test_df,phase=\"val\",test=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c2e222f5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "checkpoint.pth.tar  model_best.pth.tar\r\n"
     ]
    }
   ],
   "source": [
    "!ls runs/handle_white_neg/dropout0.3_efficientnet_b4_aug++ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0ceeea87",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Architecture: efficientnet_b4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using cache found in /root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub\n",
      "/root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub/PyTorch/Classification/ConvNets/image_classification/models/common.py:13: UserWarning: pytorch_quantization module not found, quantization will not be available\n",
      "  warnings.warn(\n",
      "/root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub/PyTorch/Classification/ConvNets/image_classification/models/efficientnet.py:17: UserWarning: pytorch_quantization module not found, quantization will not be available\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "EfficientNet(\n",
       "  (stem): Sequential(\n",
       "    (conv): Conv2d(3, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "    (bn): BatchNorm2d(48, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "    (activation): SiLU(inplace=True)\n",
       "  )\n",
       "  (layers): Sequential(\n",
       "    (0): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=48, bias=False)\n",
       "          (bn): BatchNorm2d(48, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=48, out_features=12, bias=True)\n",
       "          (expand): Linear(in_features=12, out_features=48, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(48, 24, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(24, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(24, 24, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=24, bias=False)\n",
       "          (bn): BatchNorm2d(24, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=24, out_features=6, bias=True)\n",
       "          (expand): Linear(in_features=6, out_features=24, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(24, 24, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(24, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (1): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(24, 144, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(144, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(144, 144, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=144, bias=False)\n",
       "          (bn): BatchNorm2d(144, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=144, out_features=6, bias=True)\n",
       "          (expand): Linear(in_features=6, out_features=144, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(144, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(32, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(32, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=192, bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=192, out_features=8, bias=True)\n",
       "          (expand): Linear(in_features=8, out_features=192, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(192, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(32, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block2): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(32, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=192, bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=192, out_features=8, bias=True)\n",
       "          (expand): Linear(in_features=8, out_features=192, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(192, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(32, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block3): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(32, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=192, bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=192, out_features=8, bias=True)\n",
       "          (expand): Linear(in_features=8, out_features=192, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(192, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(32, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (2): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(32, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(192, 192, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), groups=192, bias=False)\n",
       "          (bn): BatchNorm2d(192, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=192, out_features=8, bias=True)\n",
       "          (expand): Linear(in_features=8, out_features=192, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(192, 56, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(56, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(56, 336, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(336, 336, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=336, bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=336, out_features=14, bias=True)\n",
       "          (expand): Linear(in_features=14, out_features=336, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(336, 56, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(56, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block2): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(56, 336, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(336, 336, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=336, bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=336, out_features=14, bias=True)\n",
       "          (expand): Linear(in_features=14, out_features=336, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(336, 56, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(56, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block3): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(56, 336, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(336, 336, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=336, bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=336, out_features=14, bias=True)\n",
       "          (expand): Linear(in_features=14, out_features=336, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(336, 56, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(56, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (3): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(56, 336, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(336, 336, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=336, bias=False)\n",
       "          (bn): BatchNorm2d(336, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=336, out_features=14, bias=True)\n",
       "          (expand): Linear(in_features=14, out_features=336, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(336, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block2): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block3): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block4): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block5): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 112, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(112, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (4): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(112, 672, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(672, 672, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=672, bias=False)\n",
       "          (bn): BatchNorm2d(672, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=672, out_features=28, bias=True)\n",
       "          (expand): Linear(in_features=28, out_features=672, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(672, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block2): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block3): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block4): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block5): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(160, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (5): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(960, 960, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), groups=960, bias=False)\n",
       "          (bn): BatchNorm2d(960, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=960, out_features=40, bias=True)\n",
       "          (expand): Linear(in_features=40, out_features=960, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(960, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block2): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block3): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block4): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block5): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block6): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block7): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 272, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(272, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "    (6): Sequential(\n",
       "      (block0): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(272, 1632, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(1632, 1632, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=1632, bias=False)\n",
       "          (bn): BatchNorm2d(1632, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=1632, out_features=68, bias=True)\n",
       "          (expand): Linear(in_features=68, out_features=1632, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(1632, 448, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(448, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "      (block1): MBConvBlock(\n",
       "        (expand): Sequential(\n",
       "          (conv): Conv2d(448, 2688, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(2688, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (depsep): Sequential(\n",
       "          (conv): Conv2d(2688, 2688, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=2688, bias=False)\n",
       "          (bn): BatchNorm2d(2688, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "          (act): SiLU(inplace=True)\n",
       "        )\n",
       "        (se): SequentialSqueezeAndExcitation(\n",
       "          (squeeze): Linear(in_features=2688, out_features=112, bias=True)\n",
       "          (expand): Linear(in_features=112, out_features=2688, bias=True)\n",
       "          (activation): SiLU(inplace=True)\n",
       "          (sigmoid): Sigmoid()\n",
       "          (mul_a_quantizer): Identity()\n",
       "          (mul_b_quantizer): Identity()\n",
       "        )\n",
       "        (proj): Sequential(\n",
       "          (conv): Conv2d(2688, 448, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (bn): BatchNorm2d(448, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "        )\n",
       "        (residual_add): StochasticDepthResidual()\n",
       "        (residual_quantizer): Identity()\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (features): Sequential(\n",
       "    (conv): Conv2d(448, 1792, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "    (bn): BatchNorm2d(1792, eps=0.001, momentum=0.010000000000000009, affine=True, track_running_stats=True)\n",
       "    (activation): SiLU(inplace=True)\n",
       "  )\n",
       "  (classifier): Sequential(\n",
       "    (pooling): AdaptiveAvgPool2d(output_size=1)\n",
       "    (squeeze): Flatten()\n",
       "    (dropout): Dropout(p=0.4, inplace=False)\n",
       "    (fc): Sequential(\n",
       "      (0): Linear(in_features=1792, out_features=128, bias=True)\n",
       "      (1): Dropout(p=0.5, inplace=False)\n",
       "      (2): Linear(in_features=128, out_features=1, bias=True)\n",
       "    )\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " build_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7451cb1d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Architecture: efficientnet_b4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using cache found in /root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded Model Stored\n"
     ]
    }
   ],
   "source": [
    "run_path = \"runs/handle_white_neg/datacorrected_efficientnet_b4_aug++\"\n",
    "chk_path = f\"{run_path}/model_best.pth.tar\"\n",
    "state_dict  =torch.load(chk_path)[\"state_dict\"]\n",
    "state_dict[ \"classifier.fc.2.weight\"] = state_dict[\"classifier.fc.1.weight\"]\n",
    "state_dict[\"classifier.fc.2.bias\"] = state_dict[\"classifier.fc.1.bias\"]\n",
    "del state_dict[\"classifier.fc.1.weight\"]\n",
    "del state_dict[\"classifier.fc.1.bias\"]\n",
    "model =load_model(state_dict).to(CFG.device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "92f0a046",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "criterion = nn.BCEWithLogitsLoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d1ac4542",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Test : 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 29/29 [00:07<00:00,  3.86it/s]\n"
     ]
    }
   ],
   "source": [
    "test_data = test(model, test_loader, device=CFG.device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fbb20abf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[1.],\n",
       "        [1.],\n",
       "        [0.],\n",
       "        ...,\n",
       "        [1.],\n",
       "        [1.],\n",
       "        [1.]], device='cuda:0')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data[\"preds\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c6766d21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_id_worm</th>\n",
       "      <th>bw</th>\n",
       "      <th>fold</th>\n",
       "      <th>preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id_0005ef295aafe6acc63587db.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>id_00750938fb12fad4fd8fad24.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>id_009d7e2ef3c4f47b14c6c4c3.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>id_00e5892f475750f80a9f9ce5.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id_0128f4371f1b1a678062cff6.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1823</th>\n",
       "      <td>id_ff239c464c474c0936fd5d9c.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1824</th>\n",
       "      <td>id_ff52ef03790a72f1ca7c6837.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1825</th>\n",
       "      <td>id_ff6c6dc955223c93456ae3b9.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1826</th>\n",
       "      <td>id_ffc5b54765267f4aa7f015ca.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1827</th>\n",
       "      <td>id_fffdca893564aa86ff5be4cc.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1828 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        image_id_worm  bw  fold  preds\n",
       "0     id_0005ef295aafe6acc63587db.jpg   1   0.0    1.0\n",
       "1     id_00750938fb12fad4fd8fad24.jpg   1   0.0    1.0\n",
       "2     id_009d7e2ef3c4f47b14c6c4c3.jpg   0   0.0    0.0\n",
       "3     id_00e5892f475750f80a9f9ce5.jpg   0   0.0    0.0\n",
       "4     id_0128f4371f1b1a678062cff6.jpg   1   0.0    1.0\n",
       "...                               ...  ..   ...    ...\n",
       "1823  id_ff239c464c474c0936fd5d9c.jpg   0   0.0    0.0\n",
       "1824  id_ff52ef03790a72f1ca7c6837.jpg   1   0.0    1.0\n",
       "1825  id_ff6c6dc955223c93456ae3b9.jpg   1   0.0    1.0\n",
       "1826  id_ffc5b54765267f4aa7f015ca.jpg   1   0.0    1.0\n",
       "1827  id_fffdca893564aa86ff5be4cc.jpg   1   0.0    1.0\n",
       "\n",
       "[1828 rows x 4 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df[\"preds\"] = test_data[\"preds\"].cpu().detach().numpy()\n",
    "test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "56f081d2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_id_worm</th>\n",
       "      <th>bw</th>\n",
       "      <th>fold</th>\n",
       "      <th>preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>id_03e99321b0ef699d61d998a4.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>id_08a9c06cb7b912c6e8b07192.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>id_0cdcf54855d2007ceba043f6.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>id_0d05024d7f6dcfa40ea2c900.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>id_0f70fbfe9040967394c18441.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1766</th>\n",
       "      <td>id_f6561725e1d389a73e9e1cdf.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1771</th>\n",
       "      <td>id_f76f903e93da713c8523f4aa.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1772</th>\n",
       "      <td>id_f78276ef013b313b7bb59b33.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1774</th>\n",
       "      <td>id_f7aa8ec2a24ee73582f04287.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1789</th>\n",
       "      <td>id_fa94d831b15ec6456836203c.jpg</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>114 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        image_id_worm  bw  fold  preds\n",
       "24    id_03e99321b0ef699d61d998a4.jpg   0   0.0    1.0\n",
       "60    id_08a9c06cb7b912c6e8b07192.jpg   1   0.0    0.0\n",
       "92    id_0cdcf54855d2007ceba043f6.jpg   0   0.0    1.0\n",
       "94    id_0d05024d7f6dcfa40ea2c900.jpg   0   0.0    1.0\n",
       "115   id_0f70fbfe9040967394c18441.jpg   1   0.0    0.0\n",
       "...                               ...  ..   ...    ...\n",
       "1766  id_f6561725e1d389a73e9e1cdf.jpg   0   0.0    1.0\n",
       "1771  id_f76f903e93da713c8523f4aa.jpg   0   0.0    1.0\n",
       "1772  id_f78276ef013b313b7bb59b33.jpg   0   0.0    1.0\n",
       "1774  id_f7aa8ec2a24ee73582f04287.jpg   0   0.0    1.0\n",
       "1789  id_fa94d831b15ec6456836203c.jpg   0   0.0    1.0\n",
       "\n",
       "[114 rows x 4 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df[test_df.preds!=test_df.bw]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0e0cf8c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.937636761487965, 0.9552941176470588)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_metrics(test_df[\"preds\"].tolist(), test_df[\"bw\"].tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "1577446f",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df[test_df.preds!=test_df.bw].to_csv(\"val_filter_wrong_preds.csv\",index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "31492dcb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_id_worm</th>\n",
       "      <th>preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id_007159c1fa015ba6f394deeb.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>id_0087ad7987b3f82bc5b5c7d0.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>id_00ba116c0f45a71a7e0e652c.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>id_012f50e13a438474cc63b7b7.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>id_016185ddae947c538eac2c33.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2789</th>\n",
       "      <td>id_fec8d705a3291a144fb88ef0.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2793</th>\n",
       "      <td>id_ff1ab253ae19e92a44acba7f.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2795</th>\n",
       "      <td>id_ff3b7cb6761f3d632e1ff577.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2800</th>\n",
       "      <td>id_ffbcb27fa549278f47505515.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2801</th>\n",
       "      <td>id_ffc0e41e10b0c964d4a02811.jpg</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>865 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        image_id_worm  preds\n",
       "4     id_007159c1fa015ba6f394deeb.jpg    0.0\n",
       "5     id_0087ad7987b3f82bc5b5c7d0.jpg    0.0\n",
       "7     id_00ba116c0f45a71a7e0e652c.jpg    0.0\n",
       "16    id_012f50e13a438474cc63b7b7.jpg    0.0\n",
       "17    id_016185ddae947c538eac2c33.jpg    0.0\n",
       "...                               ...    ...\n",
       "2789  id_fec8d705a3291a144fb88ef0.jpg    0.0\n",
       "2793  id_ff1ab253ae19e92a44acba7f.jpg    0.0\n",
       "2795  id_ff3b7cb6761f3d632e1ff577.jpg    0.0\n",
       "2800  id_ffbcb27fa549278f47505515.jpg    0.0\n",
       "2801  id_ffc0e41e10b0c964d4a02811.jpg    0.0\n",
       "\n",
       "[865 rows x 2 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test_old = pd.read_csv(\"data/preds/test_filter_preds.csv\")\n",
    "# test_old[test_old.preds==0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "41a32070",
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_df.to_csv(\"data/preds/test_droupout0.3_filter_preds.csv\",index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f8c85f49",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "preds  preproc\traw\r\n"
     ]
    }
   ],
   "source": [
    "!ls data/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "041150ad",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}