{
  "id": 98087,
  "title": "Submission Error Solutions",
  "url": "/competitions/aptos2019-blindness-detection/discussion/98087",
  "author_name": "",
  "post_date": "2019-07-01T08:34:47.681079900Z",
  "votes": 7,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi all,</p>\n\n<p>I too had problems getting my submission to work (\"Submission CSV Not Found\"). I've got it working, but I haven't narrowed down which change fixed it (and don't want to waste submission attemps finding out).</p>\n\n<p>Here's a list of potential fixes. If anyone knows whether they are/not helpful, kindly comment below.</p>\n\n<ol>\n<li>Reduce batch size</li>\n<li>Remove console commands (I had for example: <code>!head -n 10 submission.csv</code> at the end to check output)</li>\n<li>Try/Catch around data preprocessing - I think there may be some all black/empty images which cause unexpected errors.</li>\n</ol>\n\n<p>It would be great if Kaggle could comment on whether there are any resource differences between a typical kernel and the one used to make the submission.</p>\n\n<p>Cheers,\nKarl</p>",
  "messages": [
    {
      "id": "565652",
      "postDate": "07/01/2019 08:34:47",
      "content": "<p>Hi all,</p>\n\n<p>I too had problems getting my submission to work (\"Submission CSV Not Found\"). I've got it working, but I haven't narrowed down which change fixed it (and don't want to waste submission attemps finding out).</p>\n\n<p>Here's a list of potential fixes. If anyone knows whether they are/not helpful, kindly comment below.</p>\n\n<ol>\n<li>Reduce batch size</li>\n<li>Remove console commands (I had for example: <code>!head -n 10 submission.csv</code> at the end to check output)</li>\n<li>Try/Catch around data preprocessing - I think there may be some all black/empty images which cause unexpected errors.</li>\n</ol>\n\n<p>It would be great if Kaggle could comment on whether there are any resource differences between a typical kernel and the one used to make the submission.</p>\n\n<p>Cheers,\nKarl</p>",
      "rawMarkdown": "Hi all,\n\nI too had problems getting my submission to work (\"Submission CSV Not Found\"). I've got it working, but I haven't narrowed down which change fixed it (and don't want to waste submission attemps finding out).\n\nHere's a list of potential fixes. If anyone knows whether they are/not helpful, kindly comment below.\n\n1. Reduce batch size\n2. Remove console commands (I had for example: `!head -n 10 submission.csv` at the end to check output)\n3. Try/Catch around data preprocessing - I think there may be some all black/empty images which cause unexpected errors.\n\nIt would be great if Kaggle could comment on whether there are any resource differences between a typical kernel and the one used to make the submission.\n\nCheers,\nKarl",
      "votes": null
    },
    {
      "id": "566245",
      "postDate": "07/02/2019 00:47:52",
      "content": "<p>Thank you! <br>\nI solved this problem with following.\n・Reduce batch size <br>\n・Remove console commands <br>\n・Change notebook-&gt;script  </p>",
      "rawMarkdown": "Thank you!  \nI solved this problem with following.\n・Reduce batch size  \n・Remove console commands  \n・Change notebook-&gt;script",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 566245,
      "author_name": "takuok",
      "author_url": "",
      "post_date": "07/02/2019 00:47:52",
      "content": "<p>Thank you! <br>\nI solved this problem with following.\n・Reduce batch size <br>\n・Remove console commands <br>\n・Change notebook-&gt;script  </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "565652": "Hi all,\n\nI too had problems getting my submission to work (\"Submission CSV Not Found\"). I've got it working, but I haven't narrowed down which change fixed it (and don't want to waste submission attemps finding out).\n\nHere's a list of potential fixes. If anyone knows whether they are/not helpful, kindly comment below.\n\n1. Reduce batch size\n2. Remove console commands (I had for example: `!head -n 10 submission.csv` at the end to check output)\n3. Try/Catch around data preprocessing - I think there may be some all black/empty images which cause unexpected errors.\n\nIt would be great if Kaggle could comment on whether there are any resource differences between a typical kernel and the one used to make the submission.\n\nCheers,\nKarl",
    "566245": "Thank you!  \nI solved this problem with following.\n・Reduce batch size  \n・Remove console commands  \n・Change notebook-&gt;script"
  },
  "source": "meta"
}