{
  "id": 139528,
  "title": "Submission Scoring Error",
  "url": "/competitions/deepfake-detection-challenge/discussion/139528",
  "author_name": "",
  "post_date": "2020-03-29T07:45:37.244667300Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi. I'm getting Submission Scoring Error again and again. Does this ring a bell with you, anyone? 3 screenshots are attached. Appreciate your input. Thanks in advance.</p>\n\n<p>submission.csv: generated\ncommit: successful (2596 sec)</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2Feed4438085bfdcb84d2028fbfcd87347%2F0001.png?generation=1585467835646910&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F164e9fd74cc238a08d0b788fb481a648%2F0002.png?generation=1585467853003090&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F1cc3f48092383c1cb6628b7c584bb11e%2F0003.png?generation=1585467880681191&amp;alt=media\" alt=\"\"></p>\n\n<p>Mats</p>",
  "messages": [
    {
      "id": "790008",
      "postDate": "03/29/2020 07:45:37",
      "content": "<p>Hi. I'm getting Submission Scoring Error again and again. Does this ring a bell with you, anyone? 3 screenshots are attached. Appreciate your input. Thanks in advance.</p>\n\n<p>submission.csv: generated\ncommit: successful (2596 sec)</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2Feed4438085bfdcb84d2028fbfcd87347%2F0001.png?generation=1585467835646910&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F164e9fd74cc238a08d0b788fb481a648%2F0002.png?generation=1585467853003090&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F1cc3f48092383c1cb6628b7c584bb11e%2F0003.png?generation=1585467880681191&amp;alt=media\" alt=\"\"></p>\n\n<p>Mats</p>",
      "rawMarkdown": "Hi. I'm getting Submission Scoring Error again and again. Does this ring a bell with you, anyone? 3 screenshots are attached. Appreciate your input. Thanks in advance.\n\nsubmission.csv: generated\ncommit: successful (2596 sec)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2Feed4438085bfdcb84d2028fbfcd87347%2F0001.png?generation=1585467835646910&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F164e9fd74cc238a08d0b788fb481a648%2F0002.png?generation=1585467853003090&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F1cc3f48092383c1cb6628b7c584bb11e%2F0003.png?generation=1585467880681191&amp;alt=media)\n\nMats",
      "votes": null
    },
    {
      "id": "790036",
      "postDate": "03/29/2020 08:15:47",
      "content": "<p>you should try to replace 0 and 1 with probabilities from last sigmoid/softmax layer. </p>",
      "rawMarkdown": "you should try to replace 0 and 1 with probabilities from last sigmoid/softmax layer.",
      "votes": null
    },
    {
      "id": "790041",
      "postDate": "03/29/2020 08:19:36",
      "content": "<p>Yuanzhe, thanks for good suggestion promptly! I will try it right away! 非常感謝 朋友!   Mats</p>",
      "rawMarkdown": "Yuanzhe, thanks for good suggestion promptly! I will try it right away! 非常感謝 朋友!   Mats",
      "votes": null
    },
    {
      "id": "790970",
      "postDate": "03/30/2020 02:00:49",
      "content": "<p>’16 decimal places’ didn’t work. I'm stacked in 'how can you generate decent CSV' competition LOL. I'm totally exhausted to generate CSV every 6 hors, last days .... Please help me anayone .....</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F4ef9f5a88ddbf8d24bb326f3a8afd0ad%2F004.png?generation=1585533627798834&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F18fd1543e450d39b1ecef93c688d44e2%2F005.png?generation=1585533647014994&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "’16 decimal places’ didn’t work. I'm stacked in 'how can you generate decent CSV' competition LOL. I'm totally exhausted to generate CSV every 6 hors, last days .... Please help me anayone .....\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F4ef9f5a88ddbf8d24bb326f3a8afd0ad%2F004.png?generation=1585533627798834&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F18fd1543e450d39b1ecef93c688d44e2%2F005.png?generation=1585533647014994&amp;alt=media)",
      "votes": null
    },
    {
      "id": "791078",
      "postDate": "03/30/2020 04:58:54",
      "content": "<p>4:40 UTC, 30th-Mar</p>\n\n<p>I took these 4 actions to overcome 'Submission Scoring Error' assuming 'csv data format' or 'process inappropriateness'. I will update what I get in 8 hours.</p>\n\n<p>[action-1] I modified my code pertaining to 'submission.csv output'. I used to do it with my own code that generate \"csv style\" document, without Pandas. so I replaced it with one that uses 'pd.read_csv' and 'to_csv'.\n[action-2] also replaced my favorite 'well trained' model (sigmoid output) with another 'young' one (softmax output).\n[action-3] I ran my notebook offline, but turned internet connection 'on'\n[action-4] testing to run notebook with GPU to see if any sign of decreasing process time, just in case</p>",
      "rawMarkdown": "4:40 UTC, 30th-Mar\n\nI took these 4 actions to overcome 'Submission Scoring Error' assuming 'csv data format' or 'process inappropriateness'. I will update what I get in 8 hours.\n\n[action-1] I modified my code pertaining to 'submission.csv output'. I used to do it with my own code that generate \"csv style\" document, without Pandas. so I replaced it with one that uses 'pd.read_csv' and 'to_csv'.\n[action-2] also replaced my favorite 'well trained' model (sigmoid output) with another 'young' one (softmax output).\n[action-3] I ran my notebook offline, but turned internet connection 'on'\n[action-4] testing to run notebook with GPU to see if any sign of decreasing process time, just in case",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 790036,
      "author_name": "yuanzhezhou",
      "author_url": "",
      "post_date": "03/29/2020 08:15:47",
      "content": "<p>you should try to replace 0 and 1 with probabilities from last sigmoid/softmax layer. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 790041,
      "author_name": "matsuryu",
      "author_url": "",
      "post_date": "03/29/2020 08:19:36",
      "content": "<p>Yuanzhe, thanks for good suggestion promptly! I will try it right away! 非常感謝 朋友!   Mats</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 790970,
      "author_name": "matsuryu",
      "author_url": "",
      "post_date": "03/30/2020 02:00:49",
      "content": "<p>’16 decimal places’ didn’t work. I'm stacked in 'how can you generate decent CSV' competition LOL. I'm totally exhausted to generate CSV every 6 hors, last days .... Please help me anayone .....</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F4ef9f5a88ddbf8d24bb326f3a8afd0ad%2F004.png?generation=1585533627798834&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F18fd1543e450d39b1ecef93c688d44e2%2F005.png?generation=1585533647014994&amp;alt=media\" alt=\"\"></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 791078,
      "author_name": "matsuryu",
      "author_url": "",
      "post_date": "03/30/2020 04:58:54",
      "content": "<p>4:40 UTC, 30th-Mar</p>\n\n<p>I took these 4 actions to overcome 'Submission Scoring Error' assuming 'csv data format' or 'process inappropriateness'. I will update what I get in 8 hours.</p>\n\n<p>[action-1] I modified my code pertaining to 'submission.csv output'. I used to do it with my own code that generate \"csv style\" document, without Pandas. so I replaced it with one that uses 'pd.read_csv' and 'to_csv'.\n[action-2] also replaced my favorite 'well trained' model (sigmoid output) with another 'young' one (softmax output).\n[action-3] I ran my notebook offline, but turned internet connection 'on'\n[action-4] testing to run notebook with GPU to see if any sign of decreasing process time, just in case</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "790008": "Hi. I'm getting Submission Scoring Error again and again. Does this ring a bell with you, anyone? 3 screenshots are attached. Appreciate your input. Thanks in advance.\n\nsubmission.csv: generated\ncommit: successful (2596 sec)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2Feed4438085bfdcb84d2028fbfcd87347%2F0001.png?generation=1585467835646910&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F164e9fd74cc238a08d0b788fb481a648%2F0002.png?generation=1585467853003090&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F1cc3f48092383c1cb6628b7c584bb11e%2F0003.png?generation=1585467880681191&amp;alt=media)\n\nMats",
    "790036": "you should try to replace 0 and 1 with probabilities from last sigmoid/softmax layer.",
    "790041": "Yuanzhe, thanks for good suggestion promptly! I will try it right away! 非常感謝 朋友!   Mats",
    "790970": "’16 decimal places’ didn’t work. I'm stacked in 'how can you generate decent CSV' competition LOL. I'm totally exhausted to generate CSV every 6 hors, last days .... Please help me anayone .....\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F4ef9f5a88ddbf8d24bb326f3a8afd0ad%2F004.png?generation=1585533627798834&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F3236291%2F18fd1543e450d39b1ecef93c688d44e2%2F005.png?generation=1585533647014994&amp;alt=media)",
    "791078": "4:40 UTC, 30th-Mar\n\nI took these 4 actions to overcome 'Submission Scoring Error' assuming 'csv data format' or 'process inappropriateness'. I will update what I get in 8 hours.\n\n[action-1] I modified my code pertaining to 'submission.csv output'. I used to do it with my own code that generate \"csv style\" document, without Pandas. so I replaced it with one that uses 'pd.read_csv' and 'to_csv'.\n[action-2] also replaced my favorite 'well trained' model (sigmoid output) with another 'young' one (softmax output).\n[action-3] I ran my notebook offline, but turned internet connection 'on'\n[action-4] testing to run notebook with GPU to see if any sign of decreasing process time, just in case"
  },
  "source": "meta"
}