{
  "id": 312629,
  "title": "Simple starter model",
  "url": "/competitions/ultra-mnist/discussion/312629",
  "author_name": "Konrad Banachewicz",
  "post_date": "2022-03-13T06:43:41.997000",
  "votes": 6,
  "comment_count": 0,
  "views": 0,
  "content": "<p>If you'd like start your ascent up the LB with something simple, this notebook might come in handy:</p>\n<p><a href=\"https://www.kaggle.com/konradb/umnist-model-train\" target=\"_blank\">https://www.kaggle.com/konradb/umnist-model-train</a></p>\n<p>Core idea is resnet50 + classification (28 classes). Lots of space to improve, but a useful starting point regardless ;-)</p>\n<p>I will post an inference kernel later. </p>",
  "messages": [
    {
      "id": 1720802,
      "postDate": "2022-03-13T06:43:41.997Z",
      "content": "<p>If you'd like start your ascent up the LB with something simple, this notebook might come in handy:</p>\n<p><a href=\"https://www.kaggle.com/konradb/umnist-model-train\" target=\"_blank\">https://www.kaggle.com/konradb/umnist-model-train</a></p>\n<p>Core idea is resnet50 + classification (28 classes). Lots of space to improve, but a useful starting point regardless ;-)</p>\n<p>I will post an inference kernel later. </p>",
      "rawMarkdown": "If you'd like start your ascent up the LB with something simple, this notebook might come in handy:\n\nhttps://www.kaggle.com/konradb/umnist-model-train\n\nCore idea is resnet50 + classification (28 classes). Lots of space to improve, but a useful starting point regardless ;-)\n\nI will post an inference kernel later. ",
      "votes": 6
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1720802": "If you'd like start your ascent up the LB with something simple, this notebook might come in handy:\n\nhttps://www.kaggle.com/konradb/umnist-model-train\n\nCore idea is resnet50 + classification (28 classes). Lots of space to improve, but a useful starting point regardless ;-)\n\nI will post an inference kernel later. "
  }
}