{
  "id": 299598,
  "title": "YOLOX Output Download Failing",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/299598",
  "author_name": "Obaid Rafiq",
  "post_date": "2022-01-09T01:41:16.760000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Since the kernels are limited to 9 hours of runtime, the training stops right after the 9 hour mark and whatever files are available should be downloadable. I can see from the logs that the output is 4.81Gb however, whenever I try to 'Download All', the download fails every time at around 2.1Gb.</p>\n<p>I tried downloading using Microsoft Edge as well, and the same issue prevails, so I'm pretty confident it's not a browser issue. Is there a way to salvage whatever I have trained? Perhaps something I can do in the future to avoid this problem and save the output somewhere else?</p>\n<p>Any help and/or suggestions will be greatly appreciated. Thank you.</p>",
  "messages": [
    {
      "id": 1643050,
      "postDate": "2022-01-09T01:41:16.760Z",
      "content": "<p>Since the kernels are limited to 9 hours of runtime, the training stops right after the 9 hour mark and whatever files are available should be downloadable. I can see from the logs that the output is 4.81Gb however, whenever I try to 'Download All', the download fails every time at around 2.1Gb.</p>\n<p>I tried downloading using Microsoft Edge as well, and the same issue prevails, so I'm pretty confident it's not a browser issue. Is there a way to salvage whatever I have trained? Perhaps something I can do in the future to avoid this problem and save the output somewhere else?</p>\n<p>Any help and/or suggestions will be greatly appreciated. Thank you.</p>",
      "rawMarkdown": "Since the kernels are limited to 9 hours of runtime, the training stops right after the 9 hour mark and whatever files are available should be downloadable. I can see from the logs that the output is 4.81Gb however, whenever I try to 'Download All', the download fails every time at around 2.1Gb.\n\nI tried downloading using Microsoft Edge as well, and the same issue prevails, so I'm pretty confident it's not a browser issue. Is there a way to salvage whatever I have trained? Perhaps something I can do in the future to avoid this problem and save the output somewhere else?\n\nAny help and/or suggestions will be greatly appreciated. Thank you."
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1643050": "Since the kernels are limited to 9 hours of runtime, the training stops right after the 9 hour mark and whatever files are available should be downloadable. I can see from the logs that the output is 4.81Gb however, whenever I try to 'Download All', the download fails every time at around 2.1Gb.\n\nI tried downloading using Microsoft Edge as well, and the same issue prevails, so I'm pretty confident it's not a browser issue. Is there a way to salvage whatever I have trained? Perhaps something I can do in the future to avoid this problem and save the output somewhere else?\n\nAny help and/or suggestions will be greatly appreciated. Thank you."
  }
}