{
  "id": 369700,
  "title": "Whether to support online inference？",
  "url": "/competitions/rsna-breast-cancer-detection/discussion/369700",
  "author_name": "Rongsheng Wang",
  "post_date": "2022-12-01T04:54:45.361000",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hello,<br>\nI want to train the model locally, Can I upload my model weights to submit for inference？👀</p>\n<p>Thank you！</p>",
  "messages": [
    {
      "id": 2051625,
      "postDate": "2022-12-01T14:28:40.867Z",
      "content": "<p><a href=\"https://www.kaggle.com/radek1\" target=\"_blank\">@radek1</a> thx！</p>",
      "rawMarkdown": "@radek1 thx！",
      "votes": 1,
      "replies": [
        {
          "id": 2051627,
          "postDate": "2022-12-01T14:29:32.343Z",
          "content": "<p>np <a href=\"https://www.kaggle.com/rongshengwang\" target=\"_blank\">@rongshengwang</a>, pleasure to help! 🙌 Best of luck in the competition!</p>",
          "rawMarkdown": "np @rongshengwang, pleasure to help! 🙌 Best of luck in the competition!"
        }
      ]
    },
    {
      "id": 2050982,
      "postDate": "2022-12-01T06:13:47.153Z",
      "content": "<p>The way to do it:</p>\n<p><code>train model locally -&gt; save weights -&gt; upload to Kaggle by creating a dataset with the weights (can be private) -&gt; attach the dataset to your notebook and run inference on Kaggle only</code></p>\n<p>Hope this helps! 🙂</p>",
      "rawMarkdown": "The way to do it:\n\n`train model locally -> save weights -> upload to Kaggle by creating a dataset with the weights (can be private) -> attach the dataset to your notebook and run inference on Kaggle only`\n\nHope this helps! 🙂",
      "votes": 2
    },
    {
      "id": 2050918,
      "postDate": "2022-12-01T04:54:45.363Z",
      "content": "<p>Hello,<br>\nI want to train the model locally, Can I upload my model weights to submit for inference？👀</p>\n<p>Thank you！</p>",
      "rawMarkdown": "Hello,\nI want to train the model locally, Can I upload my model weights to submit for inference？👀\n\nThank you！",
      "votes": 2
    },
    {
      "id": 2050956,
      "postDate": "2022-12-01T05:47:39.643Z",
      "content": "<p>Yes you can!!!</p>",
      "rawMarkdown": "Yes you can!!!"
    },
    {
      "id": 2050935,
      "postDate": "2022-12-01T05:02:54.530Z",
      "content": "<p>Yes, that's how it usually works</p>",
      "rawMarkdown": "Yes, that's how it usually works"
    }
  ],
  "comments": [
    {
      "id": 2051625,
      "author_name": "Rongsheng Wang",
      "author_url": "",
      "post_date": "2022-12-01T14:28:40.867000",
      "content": "<p><a href=\"https://www.kaggle.com/radek1\" target=\"_blank\">@radek1</a> thx！</p>",
      "votes": 1,
      "replies": [
        {
          "id": 2051627,
          "author_name": "Radek Osmulski",
          "author_url": "",
          "post_date": "2022-12-01T14:29:32.343000",
          "content": "<p>np <a href=\"https://www.kaggle.com/rongshengwang\" target=\"_blank\">@rongshengwang</a>, pleasure to help! 🙌 Best of luck in the competition!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 2050982,
      "author_name": "Radek Osmulski",
      "author_url": "",
      "post_date": "2022-12-01T06:13:47.153000",
      "content": "<p>The way to do it:</p>\n<p><code>train model locally -&gt; save weights -&gt; upload to Kaggle by creating a dataset with the weights (can be private) -&gt; attach the dataset to your notebook and run inference on Kaggle only</code></p>\n<p>Hope this helps! 🙂</p>",
      "votes": 2,
      "replies": []
    },
    {
      "id": 2050956,
      "author_name": "IMvision12",
      "author_url": "",
      "post_date": "2022-12-01T05:47:39.643000",
      "content": "<p>Yes you can!!!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2050935,
      "author_name": "Harshit Sheoran",
      "author_url": "",
      "post_date": "2022-12-01T05:02:54.530000",
      "content": "<p>Yes, that's how it usually works</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2051625": "@radek1 thx！",
    "2050982": "The way to do it:\n\n`train model locally -> save weights -> upload to Kaggle by creating a dataset with the weights (can be private) -> attach the dataset to your notebook and run inference on Kaggle only`\n\nHope this helps! 🙂",
    "2050918": "Hello,\nI want to train the model locally, Can I upload my model weights to submit for inference？👀\n\nThank you！",
    "2050956": "Yes you can!!!",
    "2050935": "Yes, that's how it usually works"
  }
}