{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import torch\n",
    "import numpy as np\n",
    "from torchvision import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_path = '/ssd/Datasets/I2E-ImageNet/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class I2E_NpzFolder(datasets.DatasetFolder):\n",
    "    def __init__(self, root, loader=None, extensions=['npz'], transform=None, target_transform=None, is_valid_file=None, allow_empty=False):\n",
    "        super(I2E_NpzFolder, self).__init__(root, loader, extensions, transform, target_transform, is_valid_file, allow_empty)\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        path, target = self.samples[index]\n",
    "        sample = torch.from_numpy(np.load(path)['arr_0']).float()\n",
    "        if self.transform is not None:\n",
    "            sample = self.transform(sample)\n",
    "        if self.target_transform is not None:\n",
    "            target = self.target_transform(target)\n",
    "\n",
    "        return sample, target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "len(train_dataset): 1281167, len(val_dataset): 50000\n"
     ]
    }
   ],
   "source": [
    "train_dataset = I2E_NpzFolder(root=os.path.join(dataset_path, 'train'))\n",
    "val_dataset = I2E_NpzFolder(root=os.path.join(dataset_path, 'val'))\n",
    "print(f'len(train_dataset): {len(train_dataset)}, len(val_dataset): {len(val_dataset)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "img.shape: torch.Size([8, 2, 224, 224]), label: 0\n"
     ]
    }
   ],
   "source": [
    "img, label = train_dataset[0]\n",
    "print(f'img.shape: {img.shape}, label: {label}')        # [T=8, p=2, H, W]"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pytorch291",
   "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": "3.11.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}