{
  "id": 189464,
  "title": "Error : Your notebook tried to allocate more memory than is available. It has restarted.",
  "url": "/competitions/riiid-test-answer-prediction/discussion/189464",
  "author_name": "",
  "post_date": "2020-10-07T17:11:16.153500300Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>I am getting error while loading the files what are the solution to this ? Apart setting row numbers.</p>",
  "messages": [
    {
      "id": "1041287",
      "postDate": "10/07/2020 17:11:16",
      "content": "<p>I am getting error while loading the files what are the solution to this ? Apart setting row numbers.</p>",
      "rawMarkdown": "I am getting error while loading the files what are the solution to this ? Apart setting row numbers.",
      "votes": null
    },
    {
      "id": "1041351",
      "postDate": "10/07/2020 17:35:45",
      "content": "<p>In order of easiness, it first makes sense to use the data types given in the demonstration kernel, secondly you could only load columns you actually need (if you don't need all of them) and lastly you can always resort to processing the data in batches if nothing else works.</p>\n<p>Besides, just using like 10% of the train data should probably be plenty for experimenting and getting a reasonably model running, so maybe it's best to just start out with a fraction of the full data.</p>",
      "rawMarkdown": "In order of easiness, it first makes sense to use the data types given in the demonstration kernel, secondly you could only load columns you actually need (if you don't need all of them) and lastly you can always resort to processing the data in batches if nothing else works.\n\nBesides, just using like 10% of the train data should probably be plenty for experimenting and getting a reasonably model running, so maybe it's best to just start out with a fraction of the full data.",
      "votes": null
    },
    {
      "id": "1041389",
      "postDate": "10/07/2020 18:01:43",
      "content": "<p>Thank you Alex for your answer</p>",
      "rawMarkdown": "Thank you Alex for your answer",
      "votes": null
    },
    {
      "id": "1041502",
      "postDate": "10/07/2020 19:27:08",
      "content": "<p>This question was <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/188908\" target=\"_blank\">already answered</a></p>",
      "rawMarkdown": "This question was [already answered](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/188908)",
      "votes": null
    },
    {
      "id": "1042135",
      "postDate": "10/08/2020 04:54:26",
      "content": "<p>thank you sir, I didnt notice that . This way you can load the full dataset, good</p>",
      "rawMarkdown": "thank you sir, I didnt notice that . This way you can load the full dataset, good",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1041351,
      "author_name": "spacelx",
      "author_url": "",
      "post_date": "10/07/2020 17:35:45",
      "content": "<p>In order of easiness, it first makes sense to use the data types given in the demonstration kernel, secondly you could only load columns you actually need (if you don't need all of them) and lastly you can always resort to processing the data in batches if nothing else works.</p>\n<p>Besides, just using like 10% of the train data should probably be plenty for experimenting and getting a reasonably model running, so maybe it's best to just start out with a fraction of the full data.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1041389,
          "author_name": "",
          "author_url": "",
          "post_date": "10/07/2020 18:01:43",
          "content": "<p>Thank you Alex for your answer</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1041502,
      "author_name": "sirishks",
      "author_url": "",
      "post_date": "10/07/2020 19:27:08",
      "content": "<p>This question was <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/188908\" target=\"_blank\">already answered</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 1042135,
          "author_name": "",
          "author_url": "",
          "post_date": "10/08/2020 04:54:26",
          "content": "<p>thank you sir, I didnt notice that . This way you can load the full dataset, good</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1041287": "I am getting error while loading the files what are the solution to this ? Apart setting row numbers.",
    "1041351": "In order of easiness, it first makes sense to use the data types given in the demonstration kernel, secondly you could only load columns you actually need (if you don't need all of them) and lastly you can always resort to processing the data in batches if nothing else works.\n\nBesides, just using like 10% of the train data should probably be plenty for experimenting and getting a reasonably model running, so maybe it's best to just start out with a fraction of the full data.",
    "1041389": "Thank you Alex for your answer",
    "1041502": "This question was [already answered](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/188908)",
    "1042135": "thank you sir, I didnt notice that . This way you can load the full dataset, good"
  },
  "source": "meta"
}