{
  "id": 512920,
  "title": "Make failed submission not count in submission limit?",
  "url": "/competitions/rsna-2024-lumbar-spine-degenerative-classification/discussion/512920",
  "author_name": "",
  "post_date": "2024-06-17T17:14:19.687003400Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi guys, I made my first 2 submissions today and they all failed (ohno) and now I can't test my prediction script any more (I tried to predict on both train dataset and the public test_series_descriptions.csv and it works fine). I'm afriad that I will mess up tommorrow. <br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2F1247c7e78ef302d846dd1b413e043a49%2Fbrave_is1o4bTFNx.png?generation=1718644523127761&amp;alt=media\"><br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2Faf6c80320200b61a2d7dc6518839b3f3%2Fbrave_X5GujRRyKa.png?generation=1718644536193496&amp;alt=media\"></p>\n<p>I guess making these failed submission not count for limit will help a lot because we only can make 2 submissions a day.</p>\n<p>Any help/opinion will be appreciated!</p>",
  "messages": [
    {
      "id": "2876299",
      "postDate": "06/17/2024 17:14:19",
      "content": "<p>Hi guys, I made my first 2 submissions today and they all failed (ohno) and now I can't test my prediction script any more (I tried to predict on both train dataset and the public test_series_descriptions.csv and it works fine). I'm afriad that I will mess up tommorrow. <br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2F1247c7e78ef302d846dd1b413e043a49%2Fbrave_is1o4bTFNx.png?generation=1718644523127761&amp;alt=media\"><br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2Faf6c80320200b61a2d7dc6518839b3f3%2Fbrave_X5GujRRyKa.png?generation=1718644536193496&amp;alt=media\"></p>\n<p>I guess making these failed submission not count for limit will help a lot because we only can make 2 submissions a day.</p>\n<p>Any help/opinion will be appreciated!</p>",
      "rawMarkdown": "Hi guys, I made my first 2 submissions today and they all failed (ohno) and now I can't test my prediction script any more (I tried to predict on both train dataset and the public test_series_descriptions.csv and it works fine). I'm afriad that I will mess up tommorrow. \n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2F1247c7e78ef302d846dd1b413e043a49%2Fbrave_is1o4bTFNx.png?generation=1718644523127761&alt=media)\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2Faf6c80320200b61a2d7dc6518839b3f3%2Fbrave_X5GujRRyKa.png?generation=1718644536193496&alt=media)\n\nI guess making these failed submission not count for limit will help a lot because we only can make 2 submissions a day.\n\nAny help/opinion will be appreciated!",
      "votes": null
    },
    {
      "id": "2876305",
      "postDate": "06/17/2024 17:28:40",
      "content": "<p>My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.</p>",
      "rawMarkdown": "My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.",
      "votes": null
    },
    {
      "id": "2877016",
      "postDate": "06/18/2024 06:05:43",
      "content": "<blockquote>\n  <p>My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.</p>\n</blockquote>\n<p>Thanks for the suggestion, I did find several problems after run through the whole train dataset (I didn't find it first because I let it run for 30min and nothing when wrong, so I assume it is fine for the rest)<br>\nI guess in the hidden test dataset, some of the studies does not have all three orientation. After skip these studies, my submission works!</p>",
      "rawMarkdown": "> My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.\n\nThanks for the suggestion, I did find several problems after run through the whole train dataset (I didn't find it first because I let it run for 30min and nothing when wrong, so I assume it is fine for the rest)\nI guess in the hidden test dataset, some of the studies does not have all three orientation. After skip these studies, my submission works!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2876305,
      "author_name": "coderrkj",
      "author_url": "",
      "post_date": "06/17/2024 17:28:40",
      "content": "<p>My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2877016,
          "author_name": "sakurayuyuko",
          "author_url": "",
          "post_date": "06/18/2024 06:05:43",
          "content": "<blockquote>\n  <p>My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.</p>\n</blockquote>\n<p>Thanks for the suggestion, I did find several problems after run through the whole train dataset (I didn't find it first because I let it run for 30min and nothing when wrong, so I assume it is fine for the rest)<br>\nI guess in the hidden test dataset, some of the studies does not have all three orientation. After skip these studies, my submission works!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2876299": "Hi guys, I made my first 2 submissions today and they all failed (ohno) and now I can't test my prediction script any more (I tried to predict on both train dataset and the public test_series_descriptions.csv and it works fine). I'm afriad that I will mess up tommorrow. \n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2F1247c7e78ef302d846dd1b413e043a49%2Fbrave_is1o4bTFNx.png?generation=1718644523127761&alt=media)\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11331125%2Faf6c80320200b61a2d7dc6518839b3f3%2Fbrave_X5GujRRyKa.png?generation=1718644536193496&alt=media)\n\nI guess making these failed submission not count for limit will help a lot because we only can make 2 submissions a day.\n\nAny help/opinion will be appreciated!",
    "2876305": "My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.",
    "2877016": "> My suggestion is to sample some amount, say 100 - 200 studies, from train folder and test run on the submission script you used. That may help in debugging the problem.\n\nThanks for the suggestion, I did find several problems after run through the whole train dataset (I didn't find it first because I let it run for 30min and nothing when wrong, so I assume it is fine for the rest)\nI guess in the hidden test dataset, some of the studies does not have all three orientation. After skip these studies, my submission works!"
  },
  "source": "meta"
}