{
  "id": 224523,
  "title": "Update underway",
  "url": "/competitions/hubmap-kidney-segmentation/discussion/224523",
  "author_name": "",
  "post_date": "2021-03-08T22:16:56.317595200Z",
  "votes": 24,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hi all!</p>\n<p><strong>Note: The data update is complete and submissions have been re-enabled. We will be re-running all existing submissions - this will take some time and we will announce when that process is complete. You can make new submissions in the interim, and they will be scored against the new dataset.</strong></p>\n<p>We've got a new private test set incoming. Some of the old test samples will be moved to train. The process is currently underway and may take some time. Submissions are disabled while this is taking place. We will notify you when the update is complete. Thank you for your patience!</p>",
  "messages": [
    {
      "id": "1231352",
      "postDate": "03/08/2021 22:16:56",
      "content": "<p>Hi all!</p>\n<p><strong>Note: The data update is complete and submissions have been re-enabled. We will be re-running all existing submissions - this will take some time and we will announce when that process is complete. You can make new submissions in the interim, and they will be scored against the new dataset.</strong></p>\n<p>We've got a new private test set incoming. Some of the old test samples will be moved to train. The process is currently underway and may take some time. Submissions are disabled while this is taking place. We will notify you when the update is complete. Thank you for your patience!</p>",
      "rawMarkdown": "Hi all!\n\n**Note: The data update is complete and submissions have been re-enabled. We will be re-running all existing submissions - this will take some time and we will announce when that process is complete. You can make new submissions in the interim, and they will be scored against the new dataset.**\n\nWe've got a new private test set incoming. Some of the old test samples will be moved to train. The process is currently underway and may take some time. Submissions are disabled while this is taking place. We will notify you when the update is complete. Thank you for your patience!",
      "votes": null
    },
    {
      "id": "1231371",
      "postDate": "03/08/2021 22:48:19",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> How will the leaderboard be handled? Will all submission be rescored automatically against the new testset?</p>",
      "rawMarkdown": "philculliton How will the leaderboard be handled? Will all submission be rescored automatically against the new testset?",
      "votes": null
    },
    {
      "id": "1231377",
      "postDate": "03/08/2021 23:03:02",
      "content": "<p>any chance we can get our TPU/GPU weekly quota reset? Otherwise some of us will have to wait anxiously till Sunday before we can re-train our models… Thanks!!!</p>",
      "rawMarkdown": "any chance we can get our TPU/GPU weekly quota reset? Otherwise some of us will have to wait anxiously till Sunday before we can re-train our models... Thanks!!!",
      "votes": null
    },
    {
      "id": "1232601",
      "postDate": "03/09/2021 21:13:15",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/rsmits\" target=\"_blank\">@rsmits</a> - yep! We'll be rescoring all past submissions. This involves re-running all code against the new data, and will take some time. New submissions will of course be scored against the new dataset.</p>",
      "rawMarkdown": "Hi @rsmits - yep! We'll be rescoring all past submissions. This involves re-running all code against the new data, and will take some time. New submissions will of course be scored against the new dataset.",
      "votes": null
    },
    {
      "id": "1232603",
      "postDate": "03/09/2021 21:15:26",
      "content": "<p>Good question <a href=\"https://www.kaggle.com/andrasferenczi\" target=\"_blank\">@andrasferenczi</a>! We unfortunately won't be able to reset weekly quotas BUT all past submissions will be re-run automatically against the new data, and will not count against quotas. As mentioned above this will take some time - we'll let you know when the process is complete.</p>",
      "rawMarkdown": "Good question @andrasferenczi! We unfortunately won't be able to reset weekly quotas BUT all past submissions will be re-run automatically against the new data, and will not count against quotas. As mentioned above this will take some time - we'll let you know when the process is complete.",
      "votes": null
    },
    {
      "id": "1233237",
      "postDate": "03/10/2021 08:53:07",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> When you will re-rerun the submissions, what about the underlying image when we have setup <strong>\"always use the lastest environement\"</strong>? The new kaggle image comes with GDAL 3.1.4 (that also impact RasterIO), some changes in this version make TIFF file loading a bit more complex with the need to use <code>subdatasets</code> for some cases. It not a big deal to fix but for us (and for many competitors I guess), this new version will make either submission fail or return zero score.</p>",
      "rawMarkdown": "philculliton When you will re-rerun the submissions, what about the underlying image when we have setup **\"always use the lastest environement\"**? The new kaggle image comes with GDAL 3.1.4 (that also impact RasterIO), some changes in this version make TIFF file loading a bit more complex with the need to use `subdatasets` for some cases. It not a big deal to fix but for us (and for many competitors I guess), this new version will make either submission fail or return zero score.",
      "votes": null
    },
    {
      "id": "1235137",
      "postDate": "03/11/2021 21:32:21",
      "content": "<p>plz could u take a look in ths notebook:  <a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a></p>\n<p><a href=\"https://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613\" target=\"_blank\">https://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613</a></p>\n<p>I have checked many times, created many versions. It runs well and outputs the submission.csv file. </p>\n<p>But when i submit this output, it returns with the code \"Submission CSV Not Found\"  in less then 3-4 min, meaning, <br>\nit even didnt start to read the files/ run the predictions steps. </p>\n<p>I guess it is something else.  This code runs well on my computer, on kaggle notebook, but not on submission. <br>\nthankyou.</p>",
      "rawMarkdown": "plz could u take a look in ths notebook:  @philculliton\n\nhttps://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613\n\nI have checked many times, created many versions. It runs well and outputs the submission.csv file. \n\nBut when i submit this output, it returns with the code \"Submission CSV Not Found\"  in less then 3-4 min, meaning, \nit even didnt start to read the files/ run the predictions steps. \n\nI guess it is something else.  This code runs well on my computer, on kaggle notebook, but not on submission. \nthankyou.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1231371,
      "author_name": "rsmits",
      "author_url": "",
      "post_date": "03/08/2021 22:48:19",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> How will the leaderboard be handled? Will all submission be rescored automatically against the new testset?</p>",
      "votes": null,
      "replies": [
        {
          "id": 1232601,
          "author_name": "philculliton",
          "author_url": "",
          "post_date": "03/09/2021 21:13:15",
          "content": "<p>Hi <a href=\"https://www.kaggle.com/rsmits\" target=\"_blank\">@rsmits</a> - yep! We'll be rescoring all past submissions. This involves re-running all code against the new data, and will take some time. New submissions will of course be scored against the new dataset.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1231377,
      "author_name": "andrasferenczi",
      "author_url": "",
      "post_date": "03/08/2021 23:03:02",
      "content": "<p>any chance we can get our TPU/GPU weekly quota reset? Otherwise some of us will have to wait anxiously till Sunday before we can re-train our models… Thanks!!!</p>",
      "votes": null,
      "replies": [
        {
          "id": 1232603,
          "author_name": "philculliton",
          "author_url": "",
          "post_date": "03/09/2021 21:15:26",
          "content": "<p>Good question <a href=\"https://www.kaggle.com/andrasferenczi\" target=\"_blank\">@andrasferenczi</a>! We unfortunately won't be able to reset weekly quotas BUT all past submissions will be re-run automatically against the new data, and will not count against quotas. As mentioned above this will take some time - we'll let you know when the process is complete.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1233237,
      "author_name": "mpware",
      "author_url": "",
      "post_date": "03/10/2021 08:53:07",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> When you will re-rerun the submissions, what about the underlying image when we have setup <strong>\"always use the lastest environement\"</strong>? The new kaggle image comes with GDAL 3.1.4 (that also impact RasterIO), some changes in this version make TIFF file loading a bit more complex with the need to use <code>subdatasets</code> for some cases. It not a big deal to fix but for us (and for many competitors I guess), this new version will make either submission fail or return zero score.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1235137,
      "author_name": "rpsantosakaggle",
      "author_url": "",
      "post_date": "03/11/2021 21:32:21",
      "content": "<p>plz could u take a look in ths notebook:  <a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a></p>\n<p><a href=\"https://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613\" target=\"_blank\">https://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613</a></p>\n<p>I have checked many times, created many versions. It runs well and outputs the submission.csv file. </p>\n<p>But when i submit this output, it returns with the code \"Submission CSV Not Found\"  in less then 3-4 min, meaning, <br>\nit even didnt start to read the files/ run the predictions steps. </p>\n<p>I guess it is something else.  This code runs well on my computer, on kaggle notebook, but not on submission. <br>\nthankyou.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1231352": "Hi all!\n\n**Note: The data update is complete and submissions have been re-enabled. We will be re-running all existing submissions - this will take some time and we will announce when that process is complete. You can make new submissions in the interim, and they will be scored against the new dataset.**\n\nWe've got a new private test set incoming. Some of the old test samples will be moved to train. The process is currently underway and may take some time. Submissions are disabled while this is taking place. We will notify you when the update is complete. Thank you for your patience!",
    "1231371": "philculliton How will the leaderboard be handled? Will all submission be rescored automatically against the new testset?",
    "1231377": "any chance we can get our TPU/GPU weekly quota reset? Otherwise some of us will have to wait anxiously till Sunday before we can re-train our models... Thanks!!!",
    "1232601": "Hi @rsmits - yep! We'll be rescoring all past submissions. This involves re-running all code against the new data, and will take some time. New submissions will of course be scored against the new dataset.",
    "1232603": "Good question @andrasferenczi! We unfortunately won't be able to reset weekly quotas BUT all past submissions will be re-run automatically against the new data, and will not count against quotas. As mentioned above this will take some time - we'll let you know when the process is complete.",
    "1233237": "philculliton When you will re-rerun the submissions, what about the underlying image when we have setup **\"always use the lastest environement\"**? The new kaggle image comes with GDAL 3.1.4 (that also impact RasterIO), some changes in this version make TIFF file loading a bit more complex with the need to use `subdatasets` for some cases. It not a big deal to fix but for us (and for many competitors I guess), this new version will make either submission fail or return zero score.",
    "1235137": "plz could u take a look in ths notebook:  @philculliton\n\nhttps://www.kaggle.com/rpsantosakaggle/submission-in-r-back-to-back?scriptVersionId=56456613\n\nI have checked many times, created many versions. It runs well and outputs the submission.csv file. \n\nBut when i submit this output, it returns with the code \"Submission CSV Not Found\"  in less then 3-4 min, meaning, \nit even didnt start to read the files/ run the predictions steps. \n\nI guess it is something else.  This code runs well on my computer, on kaggle notebook, but not on submission. \nthankyou."
  },
  "source": "meta"
}