{
  "id": 139747,
  "title": "Always a Notebook Timeout error",
  "url": "/competitions/deepfake-detection-challenge/discussion/139747",
  "author_name": "",
  "post_date": "2020-03-30T01:18:02.268772200Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hello All,</p>\n\n<p>We are trying to make a submission since a long time but are always getting a Notebook Timeout error. Our inference script completes and generates the submission.csv in less that 50mins for 400 videos but always gives Notebook Timeout error when we submit it for public scoring. We are using mtcnn for face detection.</p>\n\n<p>Any thoughts on why this would be happening?</p>",
  "messages": [
    {
      "id": "790946",
      "postDate": "03/30/2020 01:18:02",
      "content": "<p>Hello All,</p>\n\n<p>We are trying to make a submission since a long time but are always getting a Notebook Timeout error. Our inference script completes and generates the submission.csv in less that 50mins for 400 videos but always gives Notebook Timeout error when we submit it for public scoring. We are using mtcnn for face detection.</p>\n\n<p>Any thoughts on why this would be happening?</p>",
      "rawMarkdown": "Hello All,\n\nWe are trying to make a submission since a long time but are always getting a Notebook Timeout error. Our inference script completes and generates the submission.csv in less that 50mins for 400 videos but always gives Notebook Timeout error when we submit it for public scoring. We are using mtcnn for face detection.\n\nAny thoughts on why this would be happening?",
      "votes": null
    },
    {
      "id": "791341",
      "postDate": "03/30/2020 10:41:02",
      "content": "<p>One thing you could do is check the time in your submission notebook. If the 9 hours are almost up, stop inference and write 0.5 for the remaining videos.</p>\n\n<p>(You could also write <code>0.5 * (num_videos_done / total_videos)</code> for all the rows in the submission.csv. Based on the log loss you're getting, you can work out how many videos it did before giving the timeout.)</p>",
      "rawMarkdown": "One thing you could do is check the time in your submission notebook. If the 9 hours are almost up, stop inference and write 0.5 for the remaining videos.\n\n(You could also write `0.5 * (num_videos_done / total_videos)` for all the rows in the submission.csv. Based on the log loss you're getting, you can work out how many videos it did before giving the timeout.)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 791341,
      "author_name": "humananalog",
      "author_url": "",
      "post_date": "03/30/2020 10:41:02",
      "content": "<p>One thing you could do is check the time in your submission notebook. If the 9 hours are almost up, stop inference and write 0.5 for the remaining videos.</p>\n\n<p>(You could also write <code>0.5 * (num_videos_done / total_videos)</code> for all the rows in the submission.csv. Based on the log loss you're getting, you can work out how many videos it did before giving the timeout.)</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "790946": "Hello All,\n\nWe are trying to make a submission since a long time but are always getting a Notebook Timeout error. Our inference script completes and generates the submission.csv in less that 50mins for 400 videos but always gives Notebook Timeout error when we submit it for public scoring. We are using mtcnn for face detection.\n\nAny thoughts on why this would be happening?",
    "791341": "One thing you could do is check the time in your submission notebook. If the 9 hours are almost up, stop inference and write 0.5 for the remaining videos.\n\n(You could also write `0.5 * (num_videos_done / total_videos)` for all the rows in the submission.csv. Based on the log loss you're getting, you can work out how many videos it did before giving the timeout.)"
  },
  "source": "meta"
}