{
  "id": 671915,
  "title": "Public Dataset with Train/Test/Unlabeled Data",
  "url": "/competitions/beyond-visible-spectrum-ai-for-agriculture-2026p2/discussion/671915",
  "author_name": "",
  "post_date": "2026-02-04T22:06:12.099292500Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I have been trying for the past 2 days to create a Dataset with the 120+ GBs of Data for this competition. I know the limit in Kaggle is 200GBs but regardless what mechanism I use, kaggle cli/api/mcp and GUI, the dataset creation either fails or stays in limbo after uploading the .tar.gz file. I someone already has a private and complete dataset and its willing to share, that would be fantastic. Otherwise, given the size of this data and the problems on the kaggle side, I would encourage the host (robeson) to try and create a complete dataset at the competition level we all could use. Thanks in advance !</p>",
  "messages": [
    {
      "id": "3401939",
      "postDate": "02/04/2026 22:06:12",
      "content": "<p>I have been trying for the past 2 days to create a Dataset with the 120+ GBs of Data for this competition. I know the limit in Kaggle is 200GBs but regardless what mechanism I use, kaggle cli/api/mcp and GUI, the dataset creation either fails or stays in limbo after uploading the .tar.gz file. I someone already has a private and complete dataset and its willing to share, that would be fantastic. Otherwise, given the size of this data and the problems on the kaggle side, I would encourage the host (robeson) to try and create a complete dataset at the competition level we all could use. Thanks in advance !</p>",
      "rawMarkdown": "I have been trying for the past 2 days to create a Dataset with the 120+ GBs of Data for this competition. I know the limit in Kaggle is 200GBs but regardless what mechanism I use, kaggle cli/api/mcp and GUI, the dataset creation either fails or stays in limbo after uploading the .tar.gz file. I someone already has a private and complete dataset and its willing to share, that would be fantastic. Otherwise, given the size of this data and the problems on the kaggle side, I would encourage the host (robeson) to try and create a complete dataset at the competition level we all could use. Thanks in advance !",
      "votes": null
    },
    {
      "id": "3402364",
      "postDate": "02/05/2026 21:02:28",
      "content": "<p>Since no responses … I was able to create a dataset 1.32GBs with the unlabeled data also. If anyone needs it, please ping me.</p>",
      "rawMarkdown": "Since no responses ... I was able to create a dataset 1.32GBs with the unlabeled data also. If anyone needs it, please ping me.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3402364,
      "author_name": "alejandroayestaran",
      "author_url": "",
      "post_date": "02/05/2026 21:02:28",
      "content": "<p>Since no responses … I was able to create a dataset 1.32GBs with the unlabeled data also. If anyone needs it, please ping me.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3401939": "I have been trying for the past 2 days to create a Dataset with the 120+ GBs of Data for this competition. I know the limit in Kaggle is 200GBs but regardless what mechanism I use, kaggle cli/api/mcp and GUI, the dataset creation either fails or stays in limbo after uploading the .tar.gz file. I someone already has a private and complete dataset and its willing to share, that would be fantastic. Otherwise, given the size of this data and the problems on the kaggle side, I would encourage the host (robeson) to try and create a complete dataset at the competition level we all could use. Thanks in advance !",
    "3402364": "Since no responses ... I was able to create a dataset 1.32GBs with the unlabeled data also. If anyone needs it, please ping me."
  },
  "source": "meta"
}