{
  "id": 98287,
  "title": "What is kernel competition?",
  "url": "/competitions/aptos2019-blindness-detection/discussion/98287",
  "author_name": "",
  "post_date": "2019-07-02T16:06:02.483712100Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Do you only have to training from the kernel? or is it possible to use trained weights using local and GCP, AWS?\nI want to know the exact rules</p>",
  "messages": [
    {
      "id": "566806",
      "postDate": "07/02/2019 16:06:02",
      "content": "<p>Do you only have to training from the kernel? or is it possible to use trained weights using local and GCP, AWS?\nI want to know the exact rules</p>",
      "rawMarkdown": "Do you only have to training from the kernel? or is it possible to use trained weights using local and GCP, AWS?\nI want to know the exact rules",
      "votes": null
    },
    {
      "id": "566814",
      "postDate": "07/02/2019 16:10:08",
      "content": "<p>&gt;Submissions to this competition must be made through Kernels. Your kernel will re-run automatically against an unseen test set, and needs to output a file named submission.csv. </p>\n\n<p>Yes you can upload the weights and just do inference on kernel's... (Training can be done anywhere as you wish, but you have to.upload your weights..)\n&gt;You can still train a model offline, upload it as a dataset, and use the kernel exclusively to perform inference.</p>",
      "rawMarkdown": "&gt;Submissions to this competition must be made through Kernels. Your kernel will re-run automatically against an unseen test set, and needs to output a file named submission.csv. \n\nYes you can upload the weights and just do inference on kernel's... (Training can be done anywhere as you wish, but you have to.upload your weights..)\n&gt;You can still train a model offline, upload it as a dataset, and use the kernel exclusively to perform inference.",
      "votes": null
    },
    {
      "id": "566977",
      "postDate": "07/02/2019 21:42:30",
      "content": "<p><a href=\"/yonghan\">@yonghan</a> you can find out more information about kernels-only competitions in our <a href=\"https://www.kaggle.com/docs/competitions#other-competition-types\">documentation pages</a>! There is some great information about these competitions. It is also a good idea to familiarize yourself with this <a href=\"https://www.kaggle.com/c/aptos2019-blindness-detection/rules\">specific competition's rules page</a>.</p>",
      "rawMarkdown": "yonghan you can find out more information about kernels-only competitions in our [documentation pages](https://www.kaggle.com/docs/competitions#other-competition-types)! There is some great information about these competitions. It is also a good idea to familiarize yourself with this [specific competition's rules page](https://www.kaggle.com/c/aptos2019-blindness-detection/rules).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 566814,
      "author_name": "adityaecdrid",
      "author_url": "",
      "post_date": "07/02/2019 16:10:08",
      "content": "<p>&gt;Submissions to this competition must be made through Kernels. Your kernel will re-run automatically against an unseen test set, and needs to output a file named submission.csv. </p>\n\n<p>Yes you can upload the weights and just do inference on kernel's... (Training can be done anywhere as you wish, but you have to.upload your weights..)\n&gt;You can still train a model offline, upload it as a dataset, and use the kernel exclusively to perform inference.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 566977,
      "author_name": "randseay",
      "author_url": "",
      "post_date": "07/02/2019 21:42:30",
      "content": "<p><a href=\"/yonghan\">@yonghan</a> you can find out more information about kernels-only competitions in our <a href=\"https://www.kaggle.com/docs/competitions#other-competition-types\">documentation pages</a>! There is some great information about these competitions. It is also a good idea to familiarize yourself with this <a href=\"https://www.kaggle.com/c/aptos2019-blindness-detection/rules\">specific competition's rules page</a>.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "566806": "Do you only have to training from the kernel? or is it possible to use trained weights using local and GCP, AWS?\nI want to know the exact rules",
    "566814": "&gt;Submissions to this competition must be made through Kernels. Your kernel will re-run automatically against an unseen test set, and needs to output a file named submission.csv. \n\nYes you can upload the weights and just do inference on kernel's... (Training can be done anywhere as you wish, but you have to.upload your weights..)\n&gt;You can still train a model offline, upload it as a dataset, and use the kernel exclusively to perform inference.",
    "566977": "yonghan you can find out more information about kernels-only competitions in our [documentation pages](https://www.kaggle.com/docs/competitions#other-competition-types)! There is some great information about these competitions. It is also a good idea to familiarize yourself with this [specific competition's rules page](https://www.kaggle.com/c/aptos2019-blindness-detection/rules)."
  },
  "source": "meta"
}