{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "from torchvision import transforms, models\n",
    "from PIL import Image\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#load id_2_class.json\n",
    "import json\n",
    "\n",
    "with open('id_2_class_89.json') as json_file:\n",
    "    id_2_class = json.load(json_file)\n",
    "\n",
    "#make class_2_id dict\n",
    "\n",
    "class_2_id = {}\n",
    "for key, value in id_2_class.items():\n",
    "    class_2_id[value] = key"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_transform = transforms.Compose([\n",
    "    transforms.Resize((224, 224)),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize(mean=[0.5,0.5,0.5], std=[0.5,0.5,0.5])\n",
    "])\n",
    "\n",
    "class MaxViT(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(MaxViT, self).__init__()\n",
    "        model = models.maxvit_t(weights=\"DEFAULT\")\n",
    "        num_ftrs = model.classifier[5].in_features\n",
    "        model.classifier[5] = nn.Linear(num_ftrs, len(class_2_id))\n",
    "        self.model = model\n",
    "    def forward(self, x):\n",
    "        return self.model(x)\n",
    "\n",
    "# Instantiate the model\n",
    "model = MaxViT()\n",
    "model.load_state_dict(torch.load('best_model_89.pth'))\n",
    "model.eval()\n",
    "\n",
    "def inference(image_path, CONFIDENT_THRESHOLD=None):\n",
    "    img = Image.open(image_path).convert(\"L\").convert(\"RGB\")\n",
    "    img = test_transform(img)\n",
    "    img = img.unsqueeze(0)\n",
    "\n",
    "    with torch.no_grad():\n",
    "        output = F.softmax(model(img), dim=1)\n",
    "        confidence, predicted = torch.max(output.data, 1)\n",
    "\n",
    "    if CONFIDENT_THRESHOLD is not None and confidence.item() < CONFIDENT_THRESHOLD:\n",
    "        return \"UNKNOWN_CLASS\", confidence.item()\n",
    "    \n",
    "    return id_2_class[str(predicted.item())], confidence.item()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Volume', 0.9951752424240112)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inference(\"images/7820.jpg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Volume', 0.9951752424240112)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inference(\"images/7820.jpg\", 0.9) #0.9 should be good enough"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "textgen",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}