{
  "id": 233380,
  "title": "Slow data loading",
  "url": "/competitions/hotel-id-2021-fgvc8/discussion/233380",
  "author_name": "",
  "post_date": "2021-04-19T00:10:20.592792600Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Has anyone else run into really slow dataloading/batching from disk? Using ImageFolder with Pytorch dataloader, loading 64 images takes 4 seconds. Has anyone else had this issue/found workarounds?</p>",
  "messages": [
    {
      "id": "1277571",
      "postDate": "04/19/2021 00:10:20",
      "content": "<p>Has anyone else run into really slow dataloading/batching from disk? Using ImageFolder with Pytorch dataloader, loading 64 images takes 4 seconds. Has anyone else had this issue/found workarounds?</p>",
      "rawMarkdown": "Has anyone else run into really slow dataloading/batching from disk? Using ImageFolder with Pytorch dataloader, loading 64 images takes 4 seconds. Has anyone else had this issue/found workarounds?",
      "votes": null
    },
    {
      "id": "1742596",
      "postDate": "04/02/2022 05:26:40",
      "content": "<p>Hi I am working on the newer competition and got the same problem. Do you think it is related to how I read the images? I use cv2.imread() but could have used the function in PIL or torchvision </p>",
      "rawMarkdown": "Hi I am working on the newer competition and got the same problem. Do you think it is related to how I read the images? I use cv2.imread() but could have used the function in PIL or torchvision",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1742596,
      "author_name": "jaredfeng",
      "author_url": "",
      "post_date": "04/02/2022 05:26:40",
      "content": "<p>Hi I am working on the newer competition and got the same problem. Do you think it is related to how I read the images? I use cv2.imread() but could have used the function in PIL or torchvision </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1277571": "Has anyone else run into really slow dataloading/batching from disk? Using ImageFolder with Pytorch dataloader, loading 64 images takes 4 seconds. Has anyone else had this issue/found workarounds?",
    "1742596": "Hi I am working on the newer competition and got the same problem. Do you think it is related to how I read the images? I use cv2.imread() but could have used the function in PIL or torchvision"
  },
  "source": "meta"
}