{
  "id": 37966,
  "title": "Does using a cloud computing service is allowed?",
  "url": "/competitions/carvana-image-masking-challenge/discussion/37966",
  "author_name": "",
  "post_date": "2017-08-12T14:17:55.915896600Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hello everybody,\nI would ask something to the organizers of the competition : would using cloud computing services, like AWS, Google Cloud or FloydHub, be a contradiction to the rule that forbid us to share the data we work on with anybody else? Because as far i understand, one have to send the training data of the models to the cloud services'servers in order to get the computation done... \nI'm facing this solution because I already tried to make a submission for a kaggle competition in the past, and it seems that my computer is definitely too old and weak !\nThanks for your answers :)</p>",
  "messages": [
    {
      "id": "212730",
      "postDate": "08/12/2017 14:17:55",
      "content": "<p>Hello everybody,\nI would ask something to the organizers of the competition : would using cloud computing services, like AWS, Google Cloud or FloydHub, be a contradiction to the rule that forbid us to share the data we work on with anybody else? Because as far i understand, one have to send the training data of the models to the cloud services'servers in order to get the computation done... \nI'm facing this solution because I already tried to make a submission for a kaggle competition in the past, and it seems that my computer is definitely too old and weak !\nThanks for your answers :)</p>",
      "rawMarkdown": "Hello everybody,\nI would ask something to the organizers of the competition : would using cloud computing services, like AWS, Google Cloud or FloydHub, be a contradiction to the rule that forbid us to share the data we work on with anybody else? Because as far i understand, one have to send the training data of the models to the cloud services'servers in order to get the computation done... \nI'm facing this solution because I already tried to make a submission for a kaggle competition in the past, and it seems that my computer is definitely too old and weak !\nThanks for your answers :)",
      "votes": null
    },
    {
      "id": "212734",
      "postDate": "08/12/2017 14:31:26",
      "content": "<p>I can see i've kind of killed the English language.  I can't change it ! \nExcuse me, it's not my mother tongue ^^</p>",
      "rawMarkdown": "I can see i've kind of killed the English language.  I can't change it ! \nExcuse me, it's not my mother tongue ^^",
      "votes": null
    },
    {
      "id": "212760",
      "postDate": "08/12/2017 16:06:54",
      "content": "<p>Well Kaggle ran a competition where you HAD to use google cloud recently. I'm guessing as long as its all under your account and isn't shared with anyone, it will be fine...</p>\n\n<p>Let me know how it goes, I've been considering trying google cloud (especially their ML engine) for a while now.</p>",
      "rawMarkdown": "Well Kaggle ran a competition where you HAD to use google cloud recently. I'm guessing as long as its all under your account and isn't shared with anyone, it will be fine...\n\nLet me know how it goes, I've been considering trying google cloud (especially their ML engine) for a while now.",
      "votes": null
    },
    {
      "id": "212762",
      "postDate": "08/12/2017 16:15:12",
      "content": "<p>You're right, I remember now. But it was in a partnership with Google, which maybe make things different. Thank you anyway for your answer :)</p>\n\n<p>Actually i m planning to use FloydHub, because it seems that it's totally free until several hundred of hours of GPU use. But I heard also FloydHub is especially suited for beginners (which is my case, but maybe not yours). Anyway i ll let you know, no problem.</p>",
      "rawMarkdown": "You're right, I remember now. But it was in a partnership with Google, which maybe make things different. Thank you anyway for your answer :)\n\nActually i m planning to use FloydHub, because it seems that it's totally free until several hundred of hours of GPU use. But I heard also FloydHub is especially suited for beginners (which is my case, but maybe not yours). Anyway i ll let you know, no problem.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 212734,
      "author_name": "pseudominus",
      "author_url": "",
      "post_date": "08/12/2017 14:31:26",
      "content": "<p>I can see i've kind of killed the English language.  I can't change it ! \nExcuse me, it's not my mother tongue ^^</p>",
      "votes": null,
      "replies": [
        {
          "id": 212760,
          "author_name": "stimakov",
          "author_url": "",
          "post_date": "08/12/2017 16:06:54",
          "content": "<p>Well Kaggle ran a competition where you HAD to use google cloud recently. I'm guessing as long as its all under your account and isn't shared with anyone, it will be fine...</p>\n\n<p>Let me know how it goes, I've been considering trying google cloud (especially their ML engine) for a while now.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 212762,
          "author_name": "pseudominus",
          "author_url": "",
          "post_date": "08/12/2017 16:15:12",
          "content": "<p>You're right, I remember now. But it was in a partnership with Google, which maybe make things different. Thank you anyway for your answer :)</p>\n\n<p>Actually i m planning to use FloydHub, because it seems that it's totally free until several hundred of hours of GPU use. But I heard also FloydHub is especially suited for beginners (which is my case, but maybe not yours). Anyway i ll let you know, no problem.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "212730": "Hello everybody,\nI would ask something to the organizers of the competition : would using cloud computing services, like AWS, Google Cloud or FloydHub, be a contradiction to the rule that forbid us to share the data we work on with anybody else? Because as far i understand, one have to send the training data of the models to the cloud services'servers in order to get the computation done... \nI'm facing this solution because I already tried to make a submission for a kaggle competition in the past, and it seems that my computer is definitely too old and weak !\nThanks for your answers :)",
    "212734": "I can see i've kind of killed the English language.  I can't change it ! \nExcuse me, it's not my mother tongue ^^",
    "212760": "Well Kaggle ran a competition where you HAD to use google cloud recently. I'm guessing as long as its all under your account and isn't shared with anyone, it will be fine...\n\nLet me know how it goes, I've been considering trying google cloud (especially their ML engine) for a while now.",
    "212762": "You're right, I remember now. But it was in a partnership with Google, which maybe make things different. Thank you anyway for your answer :)\n\nActually i m planning to use FloydHub, because it seems that it's totally free until several hundred of hours of GPU use. But I heard also FloydHub is especially suited for beginners (which is my case, but maybe not yours). Anyway i ll let you know, no problem."
  },
  "source": "meta"
}