{
  "id": 87381,
  "title": "How to train model on Kernels？",
  "url": "/competitions/imet-2019-fgvc6/discussion/87381",
  "author_name": "",
  "post_date": "2019-03-31T04:21:28.801429100Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Every time I train the model on the kernel, I disconnect. Is there any way to run it offline?</p>",
  "messages": [
    {
      "id": "504168",
      "postDate": "03/31/2019 04:21:28",
      "content": "<p>Every time I train the model on the kernel, I disconnect. Is there any way to run it offline?</p>",
      "rawMarkdown": "Every time I train the model on the kernel, I disconnect. Is there any way to run it offline?",
      "votes": null
    },
    {
      "id": "504248",
      "postDate": "03/31/2019 08:34:56",
      "content": "<p>Are you using Python?  </p>\n\n<p>Are you executing via notebook?</p>\n\n<p>Executing a script (instead a notebook) there is a \"commit\" button that:</p>\n\n<ul>\n<li>creates and save a new version of your code</li>\n<li>starts an execution of your code, that runs in background</li>\n</ul>\n\n<p>If you keep connected, you can see the execution running and its log.</p>\n\n<p>But once the execution has started, you can disconnet and go home.  So you don't need to keep connected during the execution.</p>\n\n<p>Next time you enter in Kaggle, you can go to your profile page &gt; Kernels &gt; and see the execution output</p>\n\n<p>I'm very new to Kaggle Kernels, perhaps it is also possible via notebook. Hope it helps</p>",
      "rawMarkdown": "Are you using Python?  \n\nAre you executing via notebook?\n\nExecuting a script (instead a notebook) there is a \"commit\" button that:\n\n  - creates and save a new version of your code\n  - starts an execution of your code, that runs in background\n\nIf you keep connected, you can see the execution running and its log.\n\nBut once the execution has started, you can disconnet and go home.  So you don't need to keep connected during the execution.\n\nNext time you enter in Kaggle, you can go to your profile page &gt; Kernels &gt; and see the execution output\n\nI'm very new to Kaggle Kernels, perhaps it is also possible via notebook. Hope it helps",
      "votes": null
    },
    {
      "id": "504272",
      "postDate": "03/31/2019 09:58:08",
      "content": "<p>OK，Thank You！</p>",
      "rawMarkdown": "OK，Thank You！",
      "votes": null
    },
    {
      "id": "519193",
      "postDate": "04/18/2019 14:23:15",
      "content": "<p>This is all correct, also you can download notebook through File-&gt;Download option and run or change it locally and then upload it to kaggle. </p>",
      "rawMarkdown": "This is all correct, also you can download notebook through File-&gt;Download option and run or change it locally and then upload it to kaggle.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 504248,
      "author_name": "virilo",
      "author_url": "",
      "post_date": "03/31/2019 08:34:56",
      "content": "<p>Are you using Python?  </p>\n\n<p>Are you executing via notebook?</p>\n\n<p>Executing a script (instead a notebook) there is a \"commit\" button that:</p>\n\n<ul>\n<li>creates and save a new version of your code</li>\n<li>starts an execution of your code, that runs in background</li>\n</ul>\n\n<p>If you keep connected, you can see the execution running and its log.</p>\n\n<p>But once the execution has started, you can disconnet and go home.  So you don't need to keep connected during the execution.</p>\n\n<p>Next time you enter in Kaggle, you can go to your profile page &gt; Kernels &gt; and see the execution output</p>\n\n<p>I'm very new to Kaggle Kernels, perhaps it is also possible via notebook. Hope it helps</p>",
      "votes": null,
      "replies": [
        {
          "id": 519193,
          "author_name": "demonplus",
          "author_url": "",
          "post_date": "04/18/2019 14:23:15",
          "content": "<p>This is all correct, also you can download notebook through File-&gt;Download option and run or change it locally and then upload it to kaggle. </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 504272,
      "author_name": "lairuf",
      "author_url": "",
      "post_date": "03/31/2019 09:58:08",
      "content": "<p>OK，Thank You！</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "504168": "Every time I train the model on the kernel, I disconnect. Is there any way to run it offline?",
    "504248": "Are you using Python?  \n\nAre you executing via notebook?\n\nExecuting a script (instead a notebook) there is a \"commit\" button that:\n\n  - creates and save a new version of your code\n  - starts an execution of your code, that runs in background\n\nIf you keep connected, you can see the execution running and its log.\n\nBut once the execution has started, you can disconnet and go home.  So you don't need to keep connected during the execution.\n\nNext time you enter in Kaggle, you can go to your profile page &gt; Kernels &gt; and see the execution output\n\nI'm very new to Kaggle Kernels, perhaps it is also possible via notebook. Hope it helps",
    "504272": "OK，Thank You！",
    "519193": "This is all correct, also you can download notebook through File-&gt;Download option and run or change it locally and then upload it to kaggle."
  },
  "source": "meta"
}