{
  "id": 99851,
  "title": "Unknown Submission Error",
  "url": "/competitions/aptos2019-blindness-detection/discussion/99851",
  "author_name": "",
  "post_date": "2019-07-14T19:03:08.127969600Z",
  "votes": -1,
  "comment_count": 9,
  "views": 0,
  "content": "<p>hi.. is any one facing this strange issue. \nI am submitting the kernel  generating the submission file . Soon after i click on submit to competition ... kernel keeps running and then ends up in  Submission Error . \nAny help  will appreciated</p>",
  "messages": [
    {
      "id": "574965",
      "postDate": "07/14/2019 19:03:08",
      "content": "<p>hi.. is any one facing this strange issue. \nI am submitting the kernel  generating the submission file . Soon after i click on submit to competition ... kernel keeps running and then ends up in  Submission Error . \nAny help  will appreciated</p>",
      "rawMarkdown": "hi.. is any one facing this strange issue. \nI am submitting the kernel  generating the submission file . Soon after i click on submit to competition ... kernel keeps running and then ends up in  Submission Error . \nAny help  will appreciated",
      "votes": null
    },
    {
      "id": "574980",
      "postDate": "07/14/2019 19:33:16",
      "content": "<p>There are no error details provided</p>",
      "rawMarkdown": "There are no error details provided",
      "votes": null
    },
    {
      "id": "574992",
      "postDate": "07/14/2019 20:05:39",
      "content": "<p>Are you hard coding something? Remember, after submission, the kernel is run on a private test set, the results of which are used to compute LB score. So, avoid hard coding anything.</p>",
      "rawMarkdown": "Are you hard coding something? Remember, after submission, the kernel is run on a private test set, the results of which are used to compute LB score. So, avoid hard coding anything.",
      "votes": null
    },
    {
      "id": "575061",
      "postDate": "07/15/2019 02:11:50",
      "content": "<p>Thanks for reply ...what kind of hardcoding and where .</p>",
      "rawMarkdown": "Thanks for reply ...what kind of hardcoding and where .",
      "votes": null
    },
    {
      "id": "575064",
      "postDate": "07/15/2019 02:14:56",
      "content": "<p>```\nimport pandas as pd\nPATH = Path('../input/aptos2019-blindness-detection')\nPATH2 = Path('../input/')\nsubmission_df=pd.read_csv(PATH2/'submission'/'submission.csv')\nsample_submission_df=pd.read_csv(PATH2/'aptos2019-blindness-detection'/'sample_submission.csv')</p>\n\n<h1>submission_df.to_csv('submission.csv',index=False)</h1>\n\n<p>sample_submission_df['diagnosis']=submission_df['diagnosis']\nsample_submission_df.to_csv('submission.csv',index=False)\n```</p>",
      "rawMarkdown": "```\nimport pandas as pd\nPATH = Path('../input/aptos2019-blindness-detection')\nPATH2 = Path('../input/')\nsubmission_df=pd.read_csv(PATH2/'submission'/'submission.csv')\nsample_submission_df=pd.read_csv(PATH2/'aptos2019-blindness-detection'/'sample_submission.csv')\n#submission_df.to_csv('submission.csv',index=False)\n\nsample_submission_df['diagnosis']=submission_df['diagnosis']\nsample_submission_df.to_csv('submission.csv',index=False)\n```",
      "votes": null
    },
    {
      "id": "575065",
      "postDate": "07/15/2019 02:15:29",
      "content": "<p>Do you see any hardcoding </p>",
      "rawMarkdown": "Do you see any hardcoding",
      "votes": null
    },
    {
      "id": "575137",
      "postDate": "07/15/2019 05:18:30",
      "content": "<p>Are you uploading a locally generated submission csv file to the kernel? If yes, then don't do that. After you submit, the kernel is run on a private test set which has ~ 12000 images. Your submission csv is only for the public test data, not for the private test data. I hope you get it.</p>",
      "rawMarkdown": "Are you uploading a locally generated submission csv file to the kernel? If yes, then don't do that. After you submit, the kernel is run on a private test set which has ~ 12000 images. Your submission csv is only for the public test data, not for the private test data. I hope you get it.",
      "votes": null
    },
    {
      "id": "575179",
      "postDate": "07/15/2019 06:25:46",
      "content": "<p>but test set images we have got from that only random 15 pct is selected ? and rest of it goes for private set. \ncan you please tell me how it works here. how does kernel predicts the out put for the private test set .does it runs our kernel and loads those same weights which we used for generating the output ?</p>",
      "rawMarkdown": "but test set images we have got from that only random 15 pct is selected ? and rest of it goes for private set. \ncan you please tell me how it works here. how does kernel predicts the out put for the private test set .does it runs our kernel and loads those same weights which we used for generating the output ?",
      "votes": null
    },
    {
      "id": "575214",
      "postDate": "07/15/2019 07:17:50",
      "content": "<p>See, when we run our kernels we have to predict on <code>test_images</code> using <code>test.csv</code> these are the ones which have been provided to use, we call it public test set.\nNow, when we commit, kaggle reruns the kernel to generate the submission.csv, now when we submit our kernel kaggle reruns the kernel on private servers where the <code>test_images</code> and <code>test.csv</code> is replaced with the ones which have ~ 12000 images i.e., private test set. The submission.csv generated from that dataset is used to calculate the LB score (about 15% of it). Got it?</p>",
      "rawMarkdown": "See, when we run our kernels we have to predict on `test_images` using `test.csv` these are the ones which have been provided to use, we call it public test set.\nNow, when we commit, kaggle reruns the kernel to generate the submission.csv, now when we submit our kernel kaggle reruns the kernel on private servers where the `test_images` and `test.csv` is replaced with the ones which have ~ 12000 images i.e., private test set. The submission.csv generated from that dataset is used to calculate the LB score (about 15% of it). Got it?",
      "votes": null
    },
    {
      "id": "613788",
      "postDate": "08/30/2019 20:15:03",
      "content": "<p>OSError: Unable to open file (unable to open file: name = '/home/ashis/input/aptos2019-blindness-detection/efficientnet-b5_imagenet_1000_notop.h5', errno = 13, error message = 'Permission denied', flags = 0, o_flags = 0)</p>\n\n<p>how to solve this type error. plz help</p>",
      "rawMarkdown": "OSError: Unable to open file (unable to open file: name = '/home/ashis/input/aptos2019-blindness-detection/efficientnet-b5_imagenet_1000_notop.h5', errno = 13, error message = 'Permission denied', flags = 0, o_flags = 0)\n\nhow to solve this type error. plz help",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 574980,
      "author_name": "jaideepvalani",
      "author_url": "",
      "post_date": "07/14/2019 19:33:16",
      "content": "<p>There are no error details provided</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 574992,
      "author_name": "rishabhiitbhu",
      "author_url": "",
      "post_date": "07/14/2019 20:05:39",
      "content": "<p>Are you hard coding something? Remember, after submission, the kernel is run on a private test set, the results of which are used to compute LB score. So, avoid hard coding anything.</p>",
      "votes": null,
      "replies": [
        {
          "id": 575061,
          "author_name": "jaideepvalani",
          "author_url": "",
          "post_date": "07/15/2019 02:11:50",
          "content": "<p>Thanks for reply ...what kind of hardcoding and where .</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 575064,
          "author_name": "jaideepvalani",
          "author_url": "",
          "post_date": "07/15/2019 02:14:56",
          "content": "<p>```\nimport pandas as pd\nPATH = Path('../input/aptos2019-blindness-detection')\nPATH2 = Path('../input/')\nsubmission_df=pd.read_csv(PATH2/'submission'/'submission.csv')\nsample_submission_df=pd.read_csv(PATH2/'aptos2019-blindness-detection'/'sample_submission.csv')</p>\n\n<h1>submission_df.to_csv('submission.csv',index=False)</h1>\n\n<p>sample_submission_df['diagnosis']=submission_df['diagnosis']\nsample_submission_df.to_csv('submission.csv',index=False)\n```</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 575065,
          "author_name": "jaideepvalani",
          "author_url": "",
          "post_date": "07/15/2019 02:15:29",
          "content": "<p>Do you see any hardcoding </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 575137,
          "author_name": "rishabhiitbhu",
          "author_url": "",
          "post_date": "07/15/2019 05:18:30",
          "content": "<p>Are you uploading a locally generated submission csv file to the kernel? If yes, then don't do that. After you submit, the kernel is run on a private test set which has ~ 12000 images. Your submission csv is only for the public test data, not for the private test data. I hope you get it.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 575179,
          "author_name": "jaideepvalani",
          "author_url": "",
          "post_date": "07/15/2019 06:25:46",
          "content": "<p>but test set images we have got from that only random 15 pct is selected ? and rest of it goes for private set. \ncan you please tell me how it works here. how does kernel predicts the out put for the private test set .does it runs our kernel and loads those same weights which we used for generating the output ?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 575214,
          "author_name": "rishabhiitbhu",
          "author_url": "",
          "post_date": "07/15/2019 07:17:50",
          "content": "<p>See, when we run our kernels we have to predict on <code>test_images</code> using <code>test.csv</code> these are the ones which have been provided to use, we call it public test set.\nNow, when we commit, kaggle reruns the kernel to generate the submission.csv, now when we submit our kernel kaggle reruns the kernel on private servers where the <code>test_images</code> and <code>test.csv</code> is replaced with the ones which have ~ 12000 images i.e., private test set. The submission.csv generated from that dataset is used to calculate the LB score (about 15% of it). Got it?</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 613788,
      "author_name": "debasis0147",
      "author_url": "",
      "post_date": "08/30/2019 20:15:03",
      "content": "<p>OSError: Unable to open file (unable to open file: name = '/home/ashis/input/aptos2019-blindness-detection/efficientnet-b5_imagenet_1000_notop.h5', errno = 13, error message = 'Permission denied', flags = 0, o_flags = 0)</p>\n\n<p>how to solve this type error. plz help</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "574965": "hi.. is any one facing this strange issue. \nI am submitting the kernel  generating the submission file . Soon after i click on submit to competition ... kernel keeps running and then ends up in  Submission Error . \nAny help  will appreciated",
    "574980": "There are no error details provided",
    "574992": "Are you hard coding something? Remember, after submission, the kernel is run on a private test set, the results of which are used to compute LB score. So, avoid hard coding anything.",
    "575061": "Thanks for reply ...what kind of hardcoding and where .",
    "575064": "```\nimport pandas as pd\nPATH = Path('../input/aptos2019-blindness-detection')\nPATH2 = Path('../input/')\nsubmission_df=pd.read_csv(PATH2/'submission'/'submission.csv')\nsample_submission_df=pd.read_csv(PATH2/'aptos2019-blindness-detection'/'sample_submission.csv')\n#submission_df.to_csv('submission.csv',index=False)\n\nsample_submission_df['diagnosis']=submission_df['diagnosis']\nsample_submission_df.to_csv('submission.csv',index=False)\n```",
    "575065": "Do you see any hardcoding",
    "575137": "Are you uploading a locally generated submission csv file to the kernel? If yes, then don't do that. After you submit, the kernel is run on a private test set which has ~ 12000 images. Your submission csv is only for the public test data, not for the private test data. I hope you get it.",
    "575179": "but test set images we have got from that only random 15 pct is selected ? and rest of it goes for private set. \ncan you please tell me how it works here. how does kernel predicts the out put for the private test set .does it runs our kernel and loads those same weights which we used for generating the output ?",
    "575214": "See, when we run our kernels we have to predict on `test_images` using `test.csv` these are the ones which have been provided to use, we call it public test set.\nNow, when we commit, kaggle reruns the kernel to generate the submission.csv, now when we submit our kernel kaggle reruns the kernel on private servers where the `test_images` and `test.csv` is replaced with the ones which have ~ 12000 images i.e., private test set. The submission.csv generated from that dataset is used to calculate the LB score (about 15% of it). Got it?",
    "613788": "OSError: Unable to open file (unable to open file: name = '/home/ashis/input/aptos2019-blindness-detection/efficientnet-b5_imagenet_1000_notop.h5', errno = 13, error message = 'Permission denied', flags = 0, o_flags = 0)\n\nhow to solve this type error. plz help"
  },
  "source": "meta"
}