{
  "id": 96125,
  "title": "Hardware Environment for this big Competition",
  "url": "/competitions/open-images-2019-object-detection/discussion/96125",
  "author_name": "",
  "post_date": "2019-06-18T06:48:02.224664700Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I only have a notebook computer. How to build an environment for these huge data training ?</p>",
  "messages": [
    {
      "id": "554899",
      "postDate": "06/18/2019 06:48:02",
      "content": "<p>I only have a notebook computer. How to build an environment for these huge data training ?</p>",
      "rawMarkdown": "I only have a notebook computer. How to build an environment for these huge data training ?",
      "votes": null
    },
    {
      "id": "554903",
      "postDate": "06/18/2019 06:58:54",
      "content": "<p>Usually they give GCP credits at some point. Or if you didn't use it use, you get 300$ GCP credits when you join google cloud</p>",
      "rawMarkdown": "Usually they give GCP credits at some point. Or if you didn't use it use, you get 300$ GCP credits when you join google cloud",
      "votes": null
    },
    {
      "id": "554921",
      "postDate": "06/18/2019 07:24:53",
      "content": "<p>Gold winner of <a href=\"https://www.kaggle.com/c/landmark-retrieval-challenge/discussion/57855\">Google Landmark Retrieval Challenge</a> a year ago, as:</p>\n\n<p>anokasTopic Author•(1st in this Competition)•a year ago•</p>\n\n<p>Good question.</p>\n\n<p>Most of the work was done on two machines: one with 4 Pascal GPUs and another dual xeon machine (52 cores/768GB RAM, although we didn't need more than 200). At times we had access to more machines temporarily (mostly CPU).</p>",
      "rawMarkdown": "Gold winner of [Google Landmark Retrieval Challenge](https://www.kaggle.com/c/landmark-retrieval-challenge/discussion/57855) a year ago, as:\n\nanokasTopic Author•(1st in this Competition)•a year ago•\n\nGood question.\n\nMost of the work was done on two machines: one with 4 Pascal GPUs and another dual xeon machine (52 cores/768GB RAM, although we didn't need more than 200). At times we had access to more machines temporarily (mostly CPU).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 554903,
      "author_name": "moshel",
      "author_url": "",
      "post_date": "06/18/2019 06:58:54",
      "content": "<p>Usually they give GCP credits at some point. Or if you didn't use it use, you get 300$ GCP credits when you join google cloud</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 554921,
      "author_name": "zzydsz",
      "author_url": "",
      "post_date": "06/18/2019 07:24:53",
      "content": "<p>Gold winner of <a href=\"https://www.kaggle.com/c/landmark-retrieval-challenge/discussion/57855\">Google Landmark Retrieval Challenge</a> a year ago, as:</p>\n\n<p>anokasTopic Author•(1st in this Competition)•a year ago•</p>\n\n<p>Good question.</p>\n\n<p>Most of the work was done on two machines: one with 4 Pascal GPUs and another dual xeon machine (52 cores/768GB RAM, although we didn't need more than 200). At times we had access to more machines temporarily (mostly CPU).</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "554899": "I only have a notebook computer. How to build an environment for these huge data training ?",
    "554903": "Usually they give GCP credits at some point. Or if you didn't use it use, you get 300$ GCP credits when you join google cloud",
    "554921": "Gold winner of [Google Landmark Retrieval Challenge](https://www.kaggle.com/c/landmark-retrieval-challenge/discussion/57855) a year ago, as:\n\nanokasTopic Author•(1st in this Competition)•a year ago•\n\nGood question.\n\nMost of the work was done on two machines: one with 4 Pascal GPUs and another dual xeon machine (52 cores/768GB RAM, although we didn't need more than 200). At times we had access to more machines temporarily (mostly CPU)."
  },
  "source": "meta"
}