{
  "id": 681843,
  "title": "Question about modifying or regenerating the provided synthetic data",
  "url": "/competitions/accident/discussion/681843",
  "author_name": "",
  "post_date": "2026-03-17T15:23:46.971952600Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi, I have a question about the synthetic data usage policy.</p>\n<p>Since the overview says the synthetic data may be used to prototype pipelines, pretrain models, or validate ideas, would it also be allowed to modify the provided synthetic data or generate additional synthetic data based on it?</p>\n<p>For example, is it acceptable to augment it, adjust labels/annotations, or create new synthetic samples for training, as long as no real test data is used and no validation/test clips are manually labeled?</p>\n<p>Thanks for clarifying.</p>",
  "messages": [
    {
      "id": "3422369",
      "postDate": "03/17/2026 15:23:46",
      "content": "<p>Hi, I have a question about the synthetic data usage policy.</p>\n<p>Since the overview says the synthetic data may be used to prototype pipelines, pretrain models, or validate ideas, would it also be allowed to modify the provided synthetic data or generate additional synthetic data based on it?</p>\n<p>For example, is it acceptable to augment it, adjust labels/annotations, or create new synthetic samples for training, as long as no real test data is used and no validation/test clips are manually labeled?</p>\n<p>Thanks for clarifying.</p>",
      "rawMarkdown": "Hi, I have a question about the synthetic data usage policy.\n\nSince the overview says the synthetic data may be used to prototype pipelines, pretrain models, or validate ideas, would it also be allowed to modify the provided synthetic data or generate additional synthetic data based on it?\n\nFor example, is it acceptable to augment it, adjust labels/annotations, or create new synthetic samples for training, as long as no real test data is used and no validation/test clips are manually labeled?\n\nThanks for clarifying.",
      "votes": null
    },
    {
      "id": "3422373",
      "postDate": "03/17/2026 15:32:06",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/mrpc2003\" target=\"_blank\">@mrpc2003</a> ,</p>\n<p>Simple answer, yes!  :]</p>\n<p>Best,\nLukas</p>",
      "rawMarkdown": "Hi @mrpc2003 ,\n\nSimple answer, yes!  :]\n\nBest,\nLukas",
      "votes": null
    },
    {
      "id": "3424968",
      "postDate": "03/20/2026 06:21:55",
      "content": "<p>Thanks for posting this, I was really curious about this question.</p>",
      "rawMarkdown": "Thanks for posting this, I was really curious about this question.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3422373,
      "author_name": "picekl",
      "author_url": "",
      "post_date": "03/17/2026 15:32:06",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/mrpc2003\" target=\"_blank\">@mrpc2003</a> ,</p>\n<p>Simple answer, yes!  :]</p>\n<p>Best,\nLukas</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3424968,
      "author_name": "ladiposamson",
      "author_url": "",
      "post_date": "03/20/2026 06:21:55",
      "content": "<p>Thanks for posting this, I was really curious about this question.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3422369": "Hi, I have a question about the synthetic data usage policy.\n\nSince the overview says the synthetic data may be used to prototype pipelines, pretrain models, or validate ideas, would it also be allowed to modify the provided synthetic data or generate additional synthetic data based on it?\n\nFor example, is it acceptable to augment it, adjust labels/annotations, or create new synthetic samples for training, as long as no real test data is used and no validation/test clips are manually labeled?\n\nThanks for clarifying.",
    "3422373": "Hi @mrpc2003 ,\n\nSimple answer, yes!  :]\n\nBest,\nLukas",
    "3424968": "Thanks for posting this, I was really curious about this question."
  },
  "source": "meta"
}