{
  "id": 92478,
  "title": "Question Regarding Final Evaluation",
  "url": "/competitions/freesound-audio-tagging-2019/discussion/92478",
  "author_name": "",
  "post_date": "2019-05-16T21:53:47.809392800Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Might be a really basic question, but is this allowed? </p>\n\n<p>1) train model or models in Kernel A and save out the weights\n2) Upload weights as a private dataset(s)\n3) build your inference only script in Kernel B and load the weights\n4) import weights from uploaded dataset and just run inference in Kernel B for final submission\n5) Refer to Kernel A if organizers want to verify later</p>",
  "messages": [
    {
      "id": "532408",
      "postDate": "05/16/2019 21:53:47",
      "content": "<p>Might be a really basic question, but is this allowed? </p>\n\n<p>1) train model or models in Kernel A and save out the weights\n2) Upload weights as a private dataset(s)\n3) build your inference only script in Kernel B and load the weights\n4) import weights from uploaded dataset and just run inference in Kernel B for final submission\n5) Refer to Kernel A if organizers want to verify later</p>",
      "rawMarkdown": "Might be a really basic question, but is this allowed? \n\n1) train model or models in Kernel A and save out the weights\n2) Upload weights as a private dataset(s)\n3) build your inference only script in Kernel B and load the weights\n4) import weights from uploaded dataset and just run inference in Kernel B for final submission\n5) Refer to Kernel A if organizers want to verify later",
      "votes": null
    },
    {
      "id": "532410",
      "postDate": "05/16/2019 22:02:23",
      "content": "<p>Yes.</p>",
      "rawMarkdown": "Yes.",
      "votes": null
    },
    {
      "id": "532418",
      "postDate": "05/16/2019 22:16:58",
      "content": "<p>Thank you!</p>",
      "rawMarkdown": "Thank you!",
      "votes": null
    },
    {
      "id": "532611",
      "postDate": "05/17/2019 11:20:04",
      "content": "<p>I have a question. \nAccording to “Refer to Kernel A if organizers want to verify later”,can I not use a local computer for training model?  Is also training kernel only?</p>",
      "rawMarkdown": "I have a question. \nAccording to “Refer to Kernel A if organizers want to verify later”,can I not use a local computer for training model?  Is also training kernel only?",
      "votes": null
    },
    {
      "id": "532615",
      "postDate": "05/17/2019 11:31:38",
      "content": "<p>Hi <a href=\"/bossimuimu\">@bossimuimu</a> , I think this answer your question: <a href=\"https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209\">https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209</a></p>",
      "rawMarkdown": "Hi @bossimuimu , I think this answer your question: https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209",
      "votes": null
    },
    {
      "id": "532616",
      "postDate": "05/17/2019 11:36:23",
      "content": "<p>Thank you. I appreciate it</p>",
      "rawMarkdown": "Thank you. I appreciate it",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 532410,
      "author_name": "plakal",
      "author_url": "",
      "post_date": "05/16/2019 22:02:23",
      "content": "<p>Yes.</p>",
      "votes": null,
      "replies": [
        {
          "id": 532418,
          "author_name": "sairam6087",
          "author_url": "",
          "post_date": "05/16/2019 22:16:58",
          "content": "<p>Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 532611,
          "author_name": "bossimuimu",
          "author_url": "",
          "post_date": "05/17/2019 11:20:04",
          "content": "<p>I have a question. \nAccording to “Refer to Kernel A if organizers want to verify later”,can I not use a local computer for training model?  Is also training kernel only?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 532615,
          "author_name": "ebouteillon",
          "author_url": "",
          "post_date": "05/17/2019 11:31:38",
          "content": "<p>Hi <a href=\"/bossimuimu\">@bossimuimu</a> , I think this answer your question: <a href=\"https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209\">https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209</a></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 532616,
          "author_name": "bossimuimu",
          "author_url": "",
          "post_date": "05/17/2019 11:36:23",
          "content": "<p>Thank you. I appreciate it</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "532408": "Might be a really basic question, but is this allowed? \n\n1) train model or models in Kernel A and save out the weights\n2) Upload weights as a private dataset(s)\n3) build your inference only script in Kernel B and load the weights\n4) import weights from uploaded dataset and just run inference in Kernel B for final submission\n5) Refer to Kernel A if organizers want to verify later",
    "532410": "Yes.",
    "532418": "Thank you!",
    "532611": "I have a question. \nAccording to “Refer to Kernel A if organizers want to verify later”,can I not use a local computer for training model?  Is also training kernel only?",
    "532615": "Hi @bossimuimu , I think this answer your question: https://www.kaggle.com/c/freesound-audio-tagging-2019/discussion/88064#510209",
    "532616": "Thank you. I appreciate it"
  },
  "source": "meta"
}