{
  "id": 183191,
  "title": "Questions regarding submission",
  "url": "/competitions/rsna-str-pulmonary-embolism-detection/discussion/183191",
  "author_name": "Juliano Batistim",
  "post_date": "2020-09-15T23:02:41.396000",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello everyone!</p>\n<p>I am a little bit confused with submission rules. Anyone can help, please?</p>\n<p>The submission rules says:</p>\n<blockquote>\n  <ul>\n  <li>External data, freely &amp; publicly available, is allowed. This includes pre-trained models.<br>\n  Submission file must be named submission.csv</li>\n  <li>This is an inference-only code competition. Your submissions will not have access to the training images, so you must train your models elsewhere and incorporate them into your submission, without reference to the folder containing the train images.</li>\n  </ul>\n</blockquote>\n<p>So I can use some notebook to train my model, but I can't submit it, that is OK. I can this notebook to train, validate, test, evaluate if everything looks good and finally save my model.<br>\nThe submission notebook should be inference-only, so I would need to use a pre-trained model, and it is indeed allowed as well. And I can use the previous saved one. But, how can I load an external pre-trained into a notebook kernel?</p>",
  "messages": [
    {
      "id": 1012091,
      "postDate": "2020-09-15T23:02:41.397Z",
      "content": "<p>Hello everyone!</p>\n<p>I am a little bit confused with submission rules. Anyone can help, please?</p>\n<p>The submission rules says:</p>\n<blockquote>\n  <ul>\n  <li>External data, freely &amp; publicly available, is allowed. This includes pre-trained models.<br>\n  Submission file must be named submission.csv</li>\n  <li>This is an inference-only code competition. Your submissions will not have access to the training images, so you must train your models elsewhere and incorporate them into your submission, without reference to the folder containing the train images.</li>\n  </ul>\n</blockquote>\n<p>So I can use some notebook to train my model, but I can't submit it, that is OK. I can this notebook to train, validate, test, evaluate if everything looks good and finally save my model.<br>\nThe submission notebook should be inference-only, so I would need to use a pre-trained model, and it is indeed allowed as well. And I can use the previous saved one. But, how can I load an external pre-trained into a notebook kernel?</p>",
      "rawMarkdown": "Hello everyone!\n\nI am a little bit confused with submission rules. Anyone can help, please?\n\nThe submission rules says:\n\n> - External data, freely & publicly available, is allowed. This includes pre-trained models.\nSubmission file must be named submission.csv\n- This is an inference-only code competition. Your submissions will not have access to the training images, so you must train your models elsewhere and incorporate them into your submission, without reference to the folder containing the train images.\n\nSo I can use some notebook to train my model, but I can't submit it, that is OK. I can this notebook to train, validate, test, evaluate if everything looks good and finally save my model.\nThe submission notebook should be inference-only, so I would need to use a pre-trained model, and it is indeed allowed as well. And I can use the previous saved one. But, how can I load an external pre-trained into a notebook kernel?",
      "votes": 1
    },
    {
      "id": 1012316,
      "postDate": "2020-09-16T02:56:39.363Z",
      "content": "<p>Nice to know, I will try that. Just thought that Add data from another notebook will load just the inputs and not the outputs. But it makes sense loads all kernel =)</p>\n<p>Thanks a lot!</p>",
      "rawMarkdown": "Nice to know, I will try that. Just thought that Add data from another notebook will load just the inputs and not the outputs. But it makes sense loads all kernel =)\n\nThanks a lot!"
    },
    {
      "id": 1012266,
      "postDate": "2020-09-16T02:15:30.500Z",
      "content": "<p>If you're using a kaggle notebook for training, then your saved model can be accessed by using that training notebook as a dataset in another notebook.</p>",
      "rawMarkdown": "If you're using a kaggle notebook for training, then your saved model can be accessed by using that training notebook as a dataset in another notebook."
    }
  ],
  "comments": [
    {
      "id": 1012316,
      "author_name": "Juliano Batistim",
      "author_url": "",
      "post_date": "2020-09-16T02:56:39.363000",
      "content": "<p>Nice to know, I will try that. Just thought that Add data from another notebook will load just the inputs and not the outputs. But it makes sense loads all kernel =)</p>\n<p>Thanks a lot!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1012266,
      "author_name": "Tim Yee",
      "author_url": "",
      "post_date": "2020-09-16T02:15:30.500000",
      "content": "<p>If you're using a kaggle notebook for training, then your saved model can be accessed by using that training notebook as a dataset in another notebook.</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1012091": "Hello everyone!\n\nI am a little bit confused with submission rules. Anyone can help, please?\n\nThe submission rules says:\n\n> - External data, freely & publicly available, is allowed. This includes pre-trained models.\nSubmission file must be named submission.csv\n- This is an inference-only code competition. Your submissions will not have access to the training images, so you must train your models elsewhere and incorporate them into your submission, without reference to the folder containing the train images.\n\nSo I can use some notebook to train my model, but I can't submit it, that is OK. I can this notebook to train, validate, test, evaluate if everything looks good and finally save my model.\nThe submission notebook should be inference-only, so I would need to use a pre-trained model, and it is indeed allowed as well. And I can use the previous saved one. But, how can I load an external pre-trained into a notebook kernel?",
    "1012316": "Nice to know, I will try that. Just thought that Add data from another notebook will load just the inputs and not the outputs. But it makes sense loads all kernel =)\n\nThanks a lot!",
    "1012266": "If you're using a kaggle notebook for training, then your saved model can be accessed by using that training notebook as a dataset in another notebook."
  }
}