{
  "id": 206842,
  "title": "Inference/Submit Kernel",
  "url": "/competitions/ranzcr-clip-catheter-line-classification/discussion/206842",
  "author_name": "",
  "post_date": "2020-12-26T19:25:23.551663400Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>I have a general question about the Code Competition (I never participated in such a competition before):</p>\n<p>Is it allowed to first train the models offline/local, and then upload the trained model to an inference kernel such that we can make the final submission? Is this legitimate?</p>\n<p>Thanks</p>",
  "messages": [
    {
      "id": "1127717",
      "postDate": "12/26/2020 19:25:23",
      "content": "<p>Hi,</p>\n<p>I have a general question about the Code Competition (I never participated in such a competition before):</p>\n<p>Is it allowed to first train the models offline/local, and then upload the trained model to an inference kernel such that we can make the final submission? Is this legitimate?</p>\n<p>Thanks</p>",
      "rawMarkdown": "Hi,\n\nI have a general question about the Code Competition (I never participated in such a competition before):\n\nIs it allowed to first train the models offline/local, and then upload the trained model to an inference kernel such that we can make the final submission? Is this legitimate?\n\nThanks",
      "votes": null
    },
    {
      "id": "1127794",
      "postDate": "12/26/2020 21:47:02",
      "content": "<p>Yes. That is how most people are doing it. No limits on training resources. </p>",
      "rawMarkdown": "Yes. That is how most people are doing it. No limits on training resources.",
      "votes": null
    },
    {
      "id": "1127803",
      "postDate": "12/26/2020 22:05:29",
      "content": "<p>Okay. Thanks for the answer 😊</p>",
      "rawMarkdown": "Okay. Thanks for the answer 😊",
      "votes": null
    },
    {
      "id": "1130081",
      "postDate": "12/28/2020 18:26:02",
      "content": "<p>You can even access the data in other cloud platforms like Google Colab, Paperspace or even locally by using the Kaggle API. You can learn more about it <a href=\"https://www.kaggle.com/docs/api\" target=\"_blank\">here</a> .</p>",
      "rawMarkdown": "You can even access the data in other cloud platforms like Google Colab, Paperspace or even locally by using the Kaggle API. You can learn more about it [here](https://www.kaggle.com/docs/api) .",
      "votes": null
    },
    {
      "id": "1130112",
      "postDate": "12/28/2020 18:41:55",
      "content": "<p>I'm working with the Kaggle API. It is somewhat annoying to 1. Save &amp; Commit your Notebook/Kernel and 2. to submit your results. Thus, it takes quite some time to submit a new approach. Anyway, thanks for the pointer.</p>",
      "rawMarkdown": "I'm working with the Kaggle API. It is somewhat annoying to 1. Save & Commit your Notebook/Kernel and 2. to submit your results. Thus, it takes quite some time to submit a new approach. Anyway, thanks for the pointer.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1127794,
      "author_name": "richardepstein",
      "author_url": "",
      "post_date": "12/26/2020 21:47:02",
      "content": "<p>Yes. That is how most people are doing it. No limits on training resources. </p>",
      "votes": null,
      "replies": [
        {
          "id": 1127803,
          "author_name": "oberfink",
          "author_url": "",
          "post_date": "12/26/2020 22:05:29",
          "content": "<p>Okay. Thanks for the answer 😊</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1130081,
      "author_name": "atharvaingle",
      "author_url": "",
      "post_date": "12/28/2020 18:26:02",
      "content": "<p>You can even access the data in other cloud platforms like Google Colab, Paperspace or even locally by using the Kaggle API. You can learn more about it <a href=\"https://www.kaggle.com/docs/api\" target=\"_blank\">here</a> .</p>",
      "votes": null,
      "replies": [
        {
          "id": 1130112,
          "author_name": "oberfink",
          "author_url": "",
          "post_date": "12/28/2020 18:41:55",
          "content": "<p>I'm working with the Kaggle API. It is somewhat annoying to 1. Save &amp; Commit your Notebook/Kernel and 2. to submit your results. Thus, it takes quite some time to submit a new approach. Anyway, thanks for the pointer.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1127717": "Hi,\n\nI have a general question about the Code Competition (I never participated in such a competition before):\n\nIs it allowed to first train the models offline/local, and then upload the trained model to an inference kernel such that we can make the final submission? Is this legitimate?\n\nThanks",
    "1127794": "Yes. That is how most people are doing it. No limits on training resources.",
    "1127803": "Okay. Thanks for the answer 😊",
    "1130081": "You can even access the data in other cloud platforms like Google Colab, Paperspace or even locally by using the Kaggle API. You can learn more about it [here](https://www.kaggle.com/docs/api) .",
    "1130112": "I'm working with the Kaggle API. It is somewhat annoying to 1. Save & Commit your Notebook/Kernel and 2. to submit your results. Thus, it takes quite some time to submit a new approach. Anyway, thanks for the pointer."
  },
  "source": "meta"
}