{
  "id": 134760,
  "title": "How Can  I Slove 'Submission Scoring Error'?",
  "url": "/competitions/bengaliai-cv19/discussion/134760",
  "author_name": "",
  "post_date": "2020-03-10T08:37:35.421057800Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>I keep failing to submit.\nInferred from the training data, it does not seem to be a memory problem because it works without error.\nI will attach my code.\n<a href=\"https://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385\">https://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385</a>\nIf there's anything wrong, let me know.</p>",
  "messages": [
    {
      "id": "767915",
      "postDate": "03/10/2020 08:37:35",
      "content": "<p>I keep failing to submit.\nInferred from the training data, it does not seem to be a memory problem because it works without error.\nI will attach my code.\n<a href=\"https://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385\">https://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385</a>\nIf there's anything wrong, let me know.</p>",
      "rawMarkdown": "I keep failing to submit.\nInferred from the training data, it does not seem to be a memory problem because it works without error.\nI will attach my code.\nhttps://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385\nIf there's anything wrong, let me know.",
      "votes": null
    },
    {
      "id": "767951",
      "postDate": "03/10/2020 09:45:02",
      "content": "<p>take a look  <a href=\"https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading\">https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading</a>\nyou're using <code>num_workers=0</code> </p>",
      "rawMarkdown": "take a look  https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading\nyou're using `num_workers=0`",
      "votes": null
    },
    {
      "id": "768112",
      "postDate": "03/10/2020 13:30:09",
      "content": "<p>Take a lookt at this <a href=\"https://www.kaggle.com/c/bengaliai-cv19/discussion/127883\">post</a>. You may be having same problem as I did.</p>",
      "rawMarkdown": "Take a lookt at this [post](https://www.kaggle.com/c/bengaliai-cv19/discussion/127883). You may be having same problem as I did.",
      "votes": null
    },
    {
      "id": "768206",
      "postDate": "03/10/2020 14:40:47",
      "content": "<p>Thank you.\nThanks to you let me know it worked out.\nBut I don't know why it works fine</p>",
      "rawMarkdown": "Thank you.\nThanks to you let me know it worked out.\nBut I don't know why it works fine",
      "votes": null
    },
    {
      "id": "768219",
      "postDate": "03/10/2020 14:55:02",
      "content": "<p>Some images in test set have really low intensity so that threshold 80 couldn't crop any useful information and leads to failure. You can also normalize before cropping the images. Both work fine for me.</p>",
      "rawMarkdown": "Some images in test set have really low intensity so that threshold 80 couldn't crop any useful information and leads to failure. You can also normalize before cropping the images. Both work fine for me.",
      "votes": null
    },
    {
      "id": "768231",
      "postDate": "03/10/2020 15:05:25",
      "content": "<p>Thank you <a href=\"/syoya1997\">@syoya1997</a> </p>",
      "rawMarkdown": "Thank you @syoya1997",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 767951,
      "author_name": "steubk",
      "author_url": "",
      "post_date": "03/10/2020 09:45:02",
      "content": "<p>take a look  <a href=\"https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading\">https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading</a>\nyou're using <code>num_workers=0</code> </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 768112,
      "author_name": "syoya1997",
      "author_url": "",
      "post_date": "03/10/2020 13:30:09",
      "content": "<p>Take a lookt at this <a href=\"https://www.kaggle.com/c/bengaliai-cv19/discussion/127883\">post</a>. You may be having same problem as I did.</p>",
      "votes": null,
      "replies": [
        {
          "id": 768206,
          "author_name": "pyy0715",
          "author_url": "",
          "post_date": "03/10/2020 14:40:47",
          "content": "<p>Thank you.\nThanks to you let me know it worked out.\nBut I don't know why it works fine</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 768219,
          "author_name": "syoya1997",
          "author_url": "",
          "post_date": "03/10/2020 14:55:02",
          "content": "<p>Some images in test set have really low intensity so that threshold 80 couldn't crop any useful information and leads to failure. You can also normalize before cropping the images. Both work fine for me.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 768231,
          "author_name": "pyy0715",
          "author_url": "",
          "post_date": "03/10/2020 15:05:25",
          "content": "<p>Thank you <a href=\"/syoya1997\">@syoya1997</a> </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "767915": "I keep failing to submit.\nInferred from the training data, it does not seem to be a memory problem because it works without error.\nI will attach my code.\nhttps://www.kaggle.com/pyy0715/bengali-pytorch-starter?scriptVersionId=29935385\nIf there's anything wrong, let me know.",
    "767951": "take a look  https://pytorch.org/docs/stable/data.html#single-and-multi-process-data-loading\nyou're using `num_workers=0`",
    "768112": "Take a lookt at this [post](https://www.kaggle.com/c/bengaliai-cv19/discussion/127883). You may be having same problem as I did.",
    "768206": "Thank you.\nThanks to you let me know it worked out.\nBut I don't know why it works fine",
    "768219": "Some images in test set have really low intensity so that threshold 80 couldn't crop any useful information and leads to failure. You can also normalize before cropping the images. Both work fine for me.",
    "768231": "Thank you @syoya1997"
  },
  "source": "meta"
}