{
  "id": 181156,
  "title": "PROVIDE BETTER RERUN ERROR STATEMENTS/LOGS.Submissions with errors are being considered within daily limits. CSV errors.",
  "url": "/competitions/landmark-recognition-2020/discussion/181156",
  "author_name": "",
  "post_date": "2020-09-07T19:58:16.960917200Z",
  "votes": 2,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi team,</p>\n<p>Currently a lot of people are struggling with rerun related submission issues. We do not even have proper logs to refer to , so that we can correct the submission. Yet, erroneous submissions are being considered within daily submission limits. Will it not be more fair, if only successful submissions are counted only for daily limits? </p>\n<p>Also, I understand the rerun code in my case is throwing some error, causing it to abort perhaps. Again , a guess only, something, I cannot trace, as the code is working fine in demo run. Condition check for training samples is there and code inside it is working fine. Appreciate, any help.</p>\n<p>if len(train_data)!=NUM_TRAINING_IMAGES:</p>\n<pre><code># prediction generating code block. This has been checked with 100 samples and it is working as expected.\n</code></pre>\n<p>else:<br>\n    sample_submission.to_csv('submission.csv',index=False)</p>\n<p><a href=\"https://www.kaggle.com/tobwey\" target=\"_blank\">@tobwey</a>  <a href=\"https://www.kaggle.com/andrefaraujo\" target=\"_blank\">@andrefaraujo</a> </p>\n<p>Are the image ids and their corresponding labels in private training set, same as public training set? </p>",
  "messages": [
    {
      "id": "1002083",
      "postDate": "09/07/2020 19:58:16",
      "content": "<p>Hi team,</p>\n<p>Currently a lot of people are struggling with rerun related submission issues. We do not even have proper logs to refer to , so that we can correct the submission. Yet, erroneous submissions are being considered within daily submission limits. Will it not be more fair, if only successful submissions are counted only for daily limits? </p>\n<p>Also, I understand the rerun code in my case is throwing some error, causing it to abort perhaps. Again , a guess only, something, I cannot trace, as the code is working fine in demo run. Condition check for training samples is there and code inside it is working fine. Appreciate, any help.</p>\n<p>if len(train_data)!=NUM_TRAINING_IMAGES:</p>\n<pre><code># prediction generating code block. This has been checked with 100 samples and it is working as expected.\n</code></pre>\n<p>else:<br>\n    sample_submission.to_csv('submission.csv',index=False)</p>\n<p><a href=\"https://www.kaggle.com/tobwey\" target=\"_blank\">@tobwey</a>  <a href=\"https://www.kaggle.com/andrefaraujo\" target=\"_blank\">@andrefaraujo</a> </p>\n<p>Are the image ids and their corresponding labels in private training set, same as public training set? </p>",
      "rawMarkdown": "Hi team,\n\nCurrently a lot of people are struggling with rerun related submission issues. We do not even have proper logs to refer to , so that we can correct the submission. Yet, erroneous submissions are being considered within daily submission limits. Will it not be more fair, if only successful submissions are counted only for daily limits? \n\nAlso, I understand the rerun code in my case is throwing some error, causing it to abort perhaps. Again , a guess only, something, I cannot trace, as the code is working fine in demo run. Condition check for training samples is there and code inside it is working fine. Appreciate, any help.\n\n\nif len(train_data)!=NUM_TRAINING_IMAGES:\n\n\n    # prediction generating code block. This has been checked with 100 samples and it is working as expected.\n\nelse:\n    sample_submission.to_csv('submission.csv',index=False)\n\n\n@tobwey  @andrefaraujo \n\nAre the image ids and their corresponding labels in private training set, same as public training set?",
      "votes": null
    },
    {
      "id": "1003078",
      "postDate": "09/08/2020 16:38:59",
      "content": "<p><a href=\"https://www.kaggle.com/dsnil87\" target=\"_blank\">@dsnil87</a> we need to count errored submissions against the daily limit to prevent the use of errored submissions to probe the private test. I know it can be frustrating to debug blindly, but encourage you to keep at it! It can help to start with a basic working submission framework and add modules of your code piece by piece until you get an error.</p>",
      "rawMarkdown": "dsnil87 we need to count errored submissions against the daily limit to prevent the use of errored submissions to probe the private test. I know it can be frustrating to debug blindly, but encourage you to keep at it! It can help to start with a basic working submission framework and add modules of your code piece by piece until you get an error.",
      "votes": null
    },
    {
      "id": "1003768",
      "postDate": "09/09/2020 08:27:30",
      "content": "<p><a href=\"https://www.kaggle.com/wcukierski\" target=\"_blank\">@wcukierski</a> I have done as us suggested.</p>\n<p>Version 10 of my notebook got submitted successfully. But Version 14 which having only training files list fetching logic different is not working.</p>\n<p>It's giving 'Submission CSV not found' error. Can someone please help me in this?</p>",
      "rawMarkdown": "wcukierski I have done as us suggested.\n\nVersion 10 of my notebook got submitted successfully. But Version 14 which having only training files list fetching logic different is not working.\n\nIt's giving 'Submission CSV not found' error. Can someone please help me in this?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1003078,
      "author_name": "wcukierski",
      "author_url": "",
      "post_date": "09/08/2020 16:38:59",
      "content": "<p><a href=\"https://www.kaggle.com/dsnil87\" target=\"_blank\">@dsnil87</a> we need to count errored submissions against the daily limit to prevent the use of errored submissions to probe the private test. I know it can be frustrating to debug blindly, but encourage you to keep at it! It can help to start with a basic working submission framework and add modules of your code piece by piece until you get an error.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1003768,
          "author_name": "rohitdeepu17",
          "author_url": "",
          "post_date": "09/09/2020 08:27:30",
          "content": "<p><a href=\"https://www.kaggle.com/wcukierski\" target=\"_blank\">@wcukierski</a> I have done as us suggested.</p>\n<p>Version 10 of my notebook got submitted successfully. But Version 14 which having only training files list fetching logic different is not working.</p>\n<p>It's giving 'Submission CSV not found' error. Can someone please help me in this?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1002083": "Hi team,\n\nCurrently a lot of people are struggling with rerun related submission issues. We do not even have proper logs to refer to , so that we can correct the submission. Yet, erroneous submissions are being considered within daily submission limits. Will it not be more fair, if only successful submissions are counted only for daily limits? \n\nAlso, I understand the rerun code in my case is throwing some error, causing it to abort perhaps. Again , a guess only, something, I cannot trace, as the code is working fine in demo run. Condition check for training samples is there and code inside it is working fine. Appreciate, any help.\n\n\nif len(train_data)!=NUM_TRAINING_IMAGES:\n\n\n    # prediction generating code block. This has been checked with 100 samples and it is working as expected.\n\nelse:\n    sample_submission.to_csv('submission.csv',index=False)\n\n\n@tobwey  @andrefaraujo \n\nAre the image ids and their corresponding labels in private training set, same as public training set?",
    "1003078": "dsnil87 we need to count errored submissions against the daily limit to prevent the use of errored submissions to probe the private test. I know it can be frustrating to debug blindly, but encourage you to keep at it! It can help to start with a basic working submission framework and add modules of your code piece by piece until you get an error.",
    "1003768": "wcukierski I have done as us suggested.\n\nVersion 10 of my notebook got submitted successfully. But Version 14 which having only training files list fetching logic different is not working.\n\nIt's giving 'Submission CSV not found' error. Can someone please help me in this?"
  },
  "source": "meta"
}