{
  "id": 666045,
  "title": "is it ok to use DINOv2 model continually pretrained on another public biomass dataset?",
  "url": "/competitions/csiro-biomass/discussion/666045",
  "author_name": "Simon Beck",
  "post_date": "2026-01-05T05:19:54.692000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>such datasets like grassclover and deep weeds i guess would be useful for DINOv2 to adapt to downstream images.</p>",
  "messages": [
    {
      "id": 3386341,
      "postDate": "2026-01-05T05:19:54.693Z",
      "content": "<p>such datasets like grassclover and deep weeds i guess would be useful for DINOv2 to adapt to downstream images.</p>",
      "rawMarkdown": "such datasets like grassclover and deep weeds i guess would be useful for DINOv2 to adapt to downstream images.",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3386341": "such datasets like grassclover and deep weeds i guess would be useful for DINOv2 to adapt to downstream images."
  }
}