{
  "id": 199986,
  "title": "Top solutions without ConvNet?",
  "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/199986",
  "author_name": "",
  "post_date": "2020-11-28T08:17:40.874414400Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi everybody!</p>\n<p>I'm curious do anybody managed to get Top100 without using of fat convolutional nets, or without using images at all? If not, what was your best scores when you tried (and later discarded) such a models?</p>",
  "messages": [
    {
      "id": "1094000",
      "postDate": "11/28/2020 08:17:40",
      "content": "<p>Hi everybody!</p>\n<p>I'm curious do anybody managed to get Top100 without using of fat convolutional nets, or without using images at all? If not, what was your best scores when you tried (and later discarded) such a models?</p>",
      "rawMarkdown": "Hi everybody!\n\nI'm curious do anybody managed to get Top100 without using of fat convolutional nets, or without using images at all? If not, what was your best scores when you tried (and later discarded) such a models?",
      "votes": null
    },
    {
      "id": "1094163",
      "postDate": "11/28/2020 11:16:02",
      "content": "<p>There was approach based on vectornet (no raster images at all) which managed to get top 10.<br>\n<a href=\"https://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711\" target=\"_blank\">https://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711</a></p>",
      "rawMarkdown": "There was approach based on vectornet (no raster images at all) which managed to get top 10.\nhttps://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711",
      "votes": null
    },
    {
      "id": "1094175",
      "postDate": "11/28/2020 11:31:20",
      "content": "<p>Thank you Alexsey, very interesting!</p>",
      "rawMarkdown": "Thank you Alexsey, very interesting!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1094163,
      "author_name": "alexpostnikov",
      "author_url": "",
      "post_date": "11/28/2020 11:16:02",
      "content": "<p>There was approach based on vectornet (no raster images at all) which managed to get top 10.<br>\n<a href=\"https://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711\" target=\"_blank\">https://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 1094175,
          "author_name": "polishch",
          "author_url": "",
          "post_date": "11/28/2020 11:31:20",
          "content": "<p>Thank you Alexsey, very interesting!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1094000": "Hi everybody!\n\nI'm curious do anybody managed to get Top100 without using of fat convolutional nets, or without using images at all? If not, what was your best scores when you tried (and later discarded) such a models?",
    "1094163": "There was approach based on vectornet (no raster images at all) which managed to get top 10.\nhttps://www.kaggle.com/c/lyft-motion-prediction-autonomous-vehicles/discussion/199711",
    "1094175": "Thank you Alexsey, very interesting!"
  },
  "source": "meta"
}