{
  "id": 301921,
  "title": "【solved-submission score error】%cd /kaggle/working before 'inference'",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/301921",
  "author_name": "",
  "post_date": "2022-01-20T05:05:05.000315200Z",
  "votes": -1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>When I submitted with my model, I encountered ‘ submission scoring error ’.The submission code is copied to ‘ YoloR [P6/W6] … one more yolo on Kaggle [INFER]’@REMEK KINAS</p>\n<pre><code>sub_df = pd.read_csv('submission.csv')\nsub_df.head()\n</code></pre>\n<p>I tried to solve it . when run the above code, I found that my model can detect the test dataset instead of Nan.</p>\n<pre><code>index                                        annotations\n0      0  0.61 474 697 23 14 0.29 1251 539 26 22 0.27 11...\n1      1  0.17 334 297 35 26 0.12 42 172 31 24 0.12 177 ...\n2      2  0.41 1250 551 25 21 0.35 169 551 38 35 0.26 23...\n</code></pre>\n<p>Is this the problem of my model or the problem of submitting part of the code?<br>\nAnyone who can help me😭</p>",
  "messages": [
    {
      "id": "1657310",
      "postDate": "01/20/2022 05:05:05",
      "content": "<p>When I submitted with my model, I encountered ‘ submission scoring error ’.The submission code is copied to ‘ YoloR [P6/W6] … one more yolo on Kaggle [INFER]’@REMEK KINAS</p>\n<pre><code>sub_df = pd.read_csv('submission.csv')\nsub_df.head()\n</code></pre>\n<p>I tried to solve it . when run the above code, I found that my model can detect the test dataset instead of Nan.</p>\n<pre><code>index                                        annotations\n0      0  0.61 474 697 23 14 0.29 1251 539 26 22 0.27 11...\n1      1  0.17 334 297 35 26 0.12 42 172 31 24 0.12 177 ...\n2      2  0.41 1250 551 25 21 0.35 169 551 38 35 0.26 23...\n</code></pre>\n<p>Is this the problem of my model or the problem of submitting part of the code?<br>\nAnyone who can help me😭</p>",
      "rawMarkdown": "When I submitted with my model, I encountered ‘ submission scoring error ’.The submission code is copied to ‘ YoloR [P6/W6] ... one more yolo on Kaggle [INFER]’@REMEK KINAS\n  \n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n  sub_df = pd.read_csv('submission.csv')\n  sub_df.head()\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n \n\n I tried to solve it . when run the above code, I found that my model can detect the test dataset instead of Nan.\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nindex                                        annotations\n0      0  0.61 474 697 23 14 0.29 1251 539 26 22 0.27 11...\n1      1  0.17 334 297 35 26 0.12 42 172 31 24 0.12 177 ...\n2      2  0.41 1250 551 25 21 0.35 169 551 38 35 0.26 23...\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nIs this the problem of my model or the problem of submitting part of the code?\nAnyone who can help me😭",
      "votes": null
    },
    {
      "id": "1657393",
      "postDate": "01/20/2022 06:37:54",
      "content": "<p>It means that your model detects objects (probably not starfish) in images provided by Kaggle API. Reason? Hard to answer without notebook … (model quality, bug in code …).</p>",
      "rawMarkdown": "It means that your model detects objects (probably not starfish) in images provided by Kaggle API. Reason? Hard to answer without notebook ... (model quality, bug in code ...).",
      "votes": null
    },
    {
      "id": "1657469",
      "postDate": "01/20/2022 07:58:55",
      "content": "<p>maybe the test data is dynamically sampled from train dataset?</p>",
      "rawMarkdown": "maybe the test data is dynamically sampled from train dataset?",
      "votes": null
    },
    {
      "id": "1657548",
      "postDate": "01/20/2022 09:27:16",
      "content": "<p>I solved the problem!!! <br>\nI went through your code several times and realized that I forgot to switch to /kaggle/working (I don't know why) when infering for test data. Now i got a nice score for my model. Thanks again !!! 😃</p>",
      "rawMarkdown": "I solved the problem!!! \nI went through your code several times and realized that I forgot to switch to /kaggle/working (I don't know why) when infering for test data. Now i got a nice score for my model. Thanks again !!! 😃",
      "votes": null
    },
    {
      "id": "1657550",
      "postDate": "01/20/2022 09:31:05",
      "content": "<p>I showed these three images of the test data, and I think there may indeed be COTS.😲</p>",
      "rawMarkdown": "I showed these three images of the test data, and I think there may indeed be COTS.😲",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1657393,
      "author_name": "remekkinas",
      "author_url": "",
      "post_date": "01/20/2022 06:37:54",
      "content": "<p>It means that your model detects objects (probably not starfish) in images provided by Kaggle API. Reason? Hard to answer without notebook … (model quality, bug in code …).</p>",
      "votes": null,
      "replies": [
        {
          "id": 1657548,
          "author_name": "yangranran",
          "author_url": "",
          "post_date": "01/20/2022 09:27:16",
          "content": "<p>I solved the problem!!! <br>\nI went through your code several times and realized that I forgot to switch to /kaggle/working (I don't know why) when infering for test data. Now i got a nice score for my model. Thanks again !!! 😃</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1657469,
      "author_name": "dragonzhang",
      "author_url": "",
      "post_date": "01/20/2022 07:58:55",
      "content": "<p>maybe the test data is dynamically sampled from train dataset?</p>",
      "votes": null,
      "replies": [
        {
          "id": 1657550,
          "author_name": "yangranran",
          "author_url": "",
          "post_date": "01/20/2022 09:31:05",
          "content": "<p>I showed these three images of the test data, and I think there may indeed be COTS.😲</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1657310": "When I submitted with my model, I encountered ‘ submission scoring error ’.The submission code is copied to ‘ YoloR [P6/W6] ... one more yolo on Kaggle [INFER]’@REMEK KINAS\n  \n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n  sub_df = pd.read_csv('submission.csv')\n  sub_df.head()\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n \n\n I tried to solve it . when run the above code, I found that my model can detect the test dataset instead of Nan.\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nindex                                        annotations\n0      0  0.61 474 697 23 14 0.29 1251 539 26 22 0.27 11...\n1      1  0.17 334 297 35 26 0.12 42 172 31 24 0.12 177 ...\n2      2  0.41 1250 551 25 21 0.35 169 551 38 35 0.26 23...\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nIs this the problem of my model or the problem of submitting part of the code?\nAnyone who can help me😭",
    "1657393": "It means that your model detects objects (probably not starfish) in images provided by Kaggle API. Reason? Hard to answer without notebook ... (model quality, bug in code ...).",
    "1657469": "maybe the test data is dynamically sampled from train dataset?",
    "1657548": "I solved the problem!!! \nI went through your code several times and realized that I forgot to switch to /kaggle/working (I don't know why) when infering for test data. Now i got a nice score for my model. Thanks again !!! 😃",
    "1657550": "I showed these three images of the test data, and I think there may indeed be COTS.😲"
  },
  "source": "meta"
}