{
  "id": 630100,
  "title": "[Info] [UPDATED] Vesuvius `npy` and `tfrecord` format",
  "url": "/competitions/vesuvius-challenge-surface-detection/discussion/630100",
  "author_name": "",
  "post_date": "2025-11-18T06:25:23.956508800Z",
  "votes": 7,
  "comment_count": 4,
  "views": 0,
  "content": "<p>The following dataset has been updated as per original updated dataset, <a href=\"https://www.kaggle.com/competitions/vesuvius-challenge-surface-detection/discussion/664184\" target=\"_blank\">discussion</a>.</p>\n<ul>\n<li><code>npy</code> format: <a href=\"https://www.kaggle.com/datasets/ipythonx/vesuvius-npy\" target=\"_blank\">https://www.kaggle.com/datasets/ipythonx/vesuvius-npy</a></li>\n<li><code>tfrecord</code> format: <a href=\"https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords\" target=\"_blank\">https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords</a></li>\n</ul>",
  "messages": [
    {
      "id": "3335689",
      "postDate": "11/18/2025 06:25:23",
      "content": "<p>The following dataset has been updated as per original updated dataset, <a href=\"https://www.kaggle.com/competitions/vesuvius-challenge-surface-detection/discussion/664184\" target=\"_blank\">discussion</a>.</p>\n<ul>\n<li><code>npy</code> format: <a href=\"https://www.kaggle.com/datasets/ipythonx/vesuvius-npy\" target=\"_blank\">https://www.kaggle.com/datasets/ipythonx/vesuvius-npy</a></li>\n<li><code>tfrecord</code> format: <a href=\"https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords\" target=\"_blank\">https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords</a></li>\n</ul>",
      "rawMarkdown": "The following dataset has been updated as per original updated dataset, [discussion](https://www.kaggle.com/competitions/vesuvius-challenge-surface-detection/discussion/664184).\n\n- `npy` format: https://www.kaggle.com/datasets/ipythonx/vesuvius-npy\n- `tfrecord` format: https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords",
      "votes": null
    },
    {
      "id": "3353164",
      "postDate": "11/29/2025 15:39:12",
      "content": "<p>thanks for sharing.</p>",
      "rawMarkdown": "thanks for sharing.",
      "votes": null
    },
    {
      "id": "3360870",
      "postDate": "12/02/2025 17:09:40",
      "content": "<p>I have also created <code>.npz</code> which is compressed version but still quite fast…\n<a href=\"https://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz\" target=\"_blank\">https://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz</a></p>",
      "rawMarkdown": "I have also created `.npz` which is compressed version but still quite fast...\nhttps://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz",
      "votes": null
    },
    {
      "id": "3360883",
      "postDate": "12/02/2025 17:21:10",
      "content": "<p>Is it faster than npy or comparable?</p>",
      "rawMarkdown": "Is it faster than npy or comparable?",
      "votes": null
    },
    {
      "id": "3361409",
      "postDate": "12/03/2025 12:15:15",
      "content": "<p>not faster, but it takes significantly less space compared to npy, for example,  if you use it on a local machine or sync to Colab</p>\n<ul>\n<li><code>npy</code> -&gt; 51GB</li>\n<li><code>npz</code> -&gt; 22GB</li>\n</ul>",
      "rawMarkdown": "not faster, but it takes significantly less space compared to npy, for example,  if you use it on a local machine or sync to Colab\n\n - `npy` -> 51GB\n - `npz` -> 22GB",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3353164,
      "author_name": "dragonzhang",
      "author_url": "",
      "post_date": "11/29/2025 15:39:12",
      "content": "<p>thanks for sharing.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3360870,
      "author_name": "jirkaborovec",
      "author_url": "",
      "post_date": "12/02/2025 17:09:40",
      "content": "<p>I have also created <code>.npz</code> which is compressed version but still quite fast…\n<a href=\"https://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz\" target=\"_blank\">https://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 3360883,
          "author_name": "ipythonx",
          "author_url": "",
          "post_date": "12/02/2025 17:21:10",
          "content": "<p>Is it faster than npy or comparable?</p>",
          "votes": null,
          "replies": [
            {
              "id": 3361409,
              "author_name": "jirkaborovec",
              "author_url": "",
              "post_date": "12/03/2025 12:15:15",
              "content": "<p>not faster, but it takes significantly less space compared to npy, for example,  if you use it on a local machine or sync to Colab</p>\n<ul>\n<li><code>npy</code> -&gt; 51GB</li>\n<li><code>npz</code> -&gt; 22GB</li>\n</ul>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3335689": "The following dataset has been updated as per original updated dataset, [discussion](https://www.kaggle.com/competitions/vesuvius-challenge-surface-detection/discussion/664184).\n\n- `npy` format: https://www.kaggle.com/datasets/ipythonx/vesuvius-npy\n- `tfrecord` format: https://www.kaggle.com/datasets/ipythonx/vesuvius-tfrecords",
    "3353164": "thanks for sharing.",
    "3360870": "I have also created `.npz` which is compressed version but still quite fast...\nhttps://www.kaggle.com/datasets/jirkaborovec/vesuvius-surface-npz",
    "3360883": "Is it faster than npy or comparable?",
    "3361409": "not faster, but it takes significantly less space compared to npy, for example,  if you use it on a local machine or sync to Colab\n\n - `npy` -> 51GB\n - `npz` -> 22GB"
  },
  "source": "meta"
}