{
  "id": 72112,
  "title": "Faster methods of loading data?",
  "url": "/competitions/quickdraw-doodle-recognition/discussion/72112",
  "author_name": "AshishSinha",
  "post_date": "2018-11-20T13:11:31.882000",
  "votes": 2,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I don't have any local resource for processing data of this size, so I am using Kaggle kernels only. Now I want to use bigger size images and some better network with more epochs, but the problem of memory and the kernel timeout is a major hurdle for me. Any clue on how I can improve my rank with this limited computation power?</p>",
  "messages": [
    {
      "id": 424640,
      "postDate": "2018-11-20T13:11:31.883Z",
      "content": "<p>I don't have any local resource for processing data of this size, so I am using Kaggle kernels only. Now I want to use bigger size images and some better network with more epochs, but the problem of memory and the kernel timeout is a major hurdle for me. Any clue on how I can improve my rank with this limited computation power?</p>",
      "rawMarkdown": "I don't have any local resource for processing data of this size, so I am using Kaggle kernels only. Now I want to use bigger size images and some better network with more epochs, but the problem of memory and the kernel timeout is a major hurdle for me. Any clue on how I can improve my rank with this limited computation power?",
      "votes": 2
    },
    {
      "id": 424865,
      "postDate": "2018-11-20T19:40:53.697Z",
      "content": "<p>For HR recruiters who want to hire data scientists, they can check \n<a href=\"https://www.interviewmocha.com/tests/data-science-analytics-test-assessment?utm_source=medium&amp;utm_medium=utkarsh-backlink\"> data science test </a> for further details.</p>",
      "rawMarkdown": "For HR recruiters who want to hire data scientists, they can check \n<a href=\"https://www.interviewmocha.com/tests/data-science-analytics-test-assessment?utm_source=medium&amp;utm_medium=utkarsh-backlink\"> data science test </a> for further details.",
      "votes": -6
    },
    {
      "id": 425411,
      "postDate": "2018-11-21T15:03:23.630Z",
      "content": "<p>You can use google cloud's free 300$ credit if you can create a new google account ;)</p>",
      "rawMarkdown": "You can use google cloud's free 300$ credit if you can create a new google account ;)"
    },
    {
      "id": 424643,
      "postDate": "2018-11-20T13:19:37.433Z",
      "content": "<p>I think google colab is more powerful than Kaggle kernels. Also, you can run their GPUs for 12 hours at a stretch.\n<a href=\"https://colab.research.google.com/\">https://colab.research.google.com/</a></p>",
      "rawMarkdown": "I think google colab is more powerful than Kaggle kernels. Also, you can run their GPUs for 12 hours at a stretch.\nhttps://colab.research.google.com/\n"
    }
  ],
  "comments": [
    {
      "id": 424865,
      "author_name": "utkarsh singh",
      "author_url": "",
      "post_date": "2018-11-20T19:40:53.697000",
      "content": "<p>For HR recruiters who want to hire data scientists, they can check \n<a href=\"https://www.interviewmocha.com/tests/data-science-analytics-test-assessment?utm_source=medium&amp;utm_medium=utkarsh-backlink\"> data science test </a> for further details.</p>",
      "votes": -6,
      "replies": []
    },
    {
      "id": 425411,
      "author_name": "remidi",
      "author_url": "",
      "post_date": "2018-11-21T15:03:23.630000",
      "content": "<p>You can use google cloud's free 300$ credit if you can create a new google account ;)</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 424643,
      "author_name": "Allen",
      "author_url": "",
      "post_date": "2018-11-20T13:19:37.433000",
      "content": "<p>I think google colab is more powerful than Kaggle kernels. Also, you can run their GPUs for 12 hours at a stretch.\n<a href=\"https://colab.research.google.com/\">https://colab.research.google.com/</a></p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "424640": "I don't have any local resource for processing data of this size, so I am using Kaggle kernels only. Now I want to use bigger size images and some better network with more epochs, but the problem of memory and the kernel timeout is a major hurdle for me. Any clue on how I can improve my rank with this limited computation power?",
    "424865": "For HR recruiters who want to hire data scientists, they can check \n<a href=\"https://www.interviewmocha.com/tests/data-science-analytics-test-assessment?utm_source=medium&amp;utm_medium=utkarsh-backlink\"> data science test </a> for further details.",
    "425411": "You can use google cloud's free 300$ credit if you can create a new google account ;)",
    "424643": "I think google colab is more powerful than Kaggle kernels. Also, you can run their GPUs for 12 hours at a stretch.\nhttps://colab.research.google.com/\n"
  }
}