{
  "id": 411863,
  "title": "Need help with a persistent submission score error",
  "url": "/competitions/birdclef-2023/discussion/411863",
  "author_name": "",
  "post_date": "2023-05-21T10:44:49.039869500Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hello everybody,</p>\n<p>I have been receiving a submission score error and I can't figure out what is causing it.. I am following more or less the same steps as the \"Inferring Birds with Kaggle Models\" notebook. My code outputs a 60x265 dataset of the same value types for a single file (10mn) and a 120x265 if it takes two files so it should be robust to the files replacement when scoring. I checked the value types \"np.float32\" and index = false when exporting the dataset. Does anybody have an idea what my problem could be ?<br>\nTwo things I am doing differently are that I assign 1.0 and 0.0 to some values but I don't see why it could cause a problem and I only process 70% of the test files with my model because it is a bit slow anf for the remaining 30% I assign 4.798188e-08 to all the bird species to have a (191x60) x 265 dataset. I would be very happy if somebody lent me hand.. Also does somebody know if it possible to contact kaggle support or the organizers for this kind of issue ? Thanks a lot and have a good day :)</p>\n<p>This is what I get if I process the given file \"soundscape_29201\" 191 times if it can help</p>",
  "messages": [
    {
      "id": "2267922",
      "postDate": "05/21/2023 10:44:49",
      "content": "<p>Hello everybody,</p>\n<p>I have been receiving a submission score error and I can't figure out what is causing it.. I am following more or less the same steps as the \"Inferring Birds with Kaggle Models\" notebook. My code outputs a 60x265 dataset of the same value types for a single file (10mn) and a 120x265 if it takes two files so it should be robust to the files replacement when scoring. I checked the value types \"np.float32\" and index = false when exporting the dataset. Does anybody have an idea what my problem could be ?<br>\nTwo things I am doing differently are that I assign 1.0 and 0.0 to some values but I don't see why it could cause a problem and I only process 70% of the test files with my model because it is a bit slow anf for the remaining 30% I assign 4.798188e-08 to all the bird species to have a (191x60) x 265 dataset. I would be very happy if somebody lent me hand.. Also does somebody know if it possible to contact kaggle support or the organizers for this kind of issue ? Thanks a lot and have a good day :)</p>\n<p>This is what I get if I process the given file \"soundscape_29201\" 191 times if it can help</p>",
      "rawMarkdown": "Hello everybody,\n\nI have been receiving a submission score error and I can't figure out what is causing it.. I am following more or less the same steps as the \"Inferring Birds with Kaggle Models\" notebook. My code outputs a 60x265 dataset of the same value types for a single file (10mn) and a 120x265 if it takes two files so it should be robust to the files replacement when scoring. I checked the value types \"np.float32\" and index = false when exporting the dataset. Does anybody have an idea what my problem could be ?\nTwo things I am doing differently are that I assign 1.0 and 0.0 to some values but I don't see why it could cause a problem and I only process 70% of the test files with my model because it is a bit slow anf for the remaining 30% I assign 4.798188e-08 to all the bird species to have a (191x60) x 265 dataset. I would be very happy if somebody lent me hand.. Also does somebody know if it possible to contact kaggle support or the organizers for this kind of issue ? Thanks a lot and have a good day :)\n\nThis is what I get if I process the given file \"soundscape_29201\" 191 times if it can help",
      "votes": null
    },
    {
      "id": "2267945",
      "postDate": "05/21/2023 10:53:26",
      "content": "<p>I don't know much, but surely can tell that assigning 1.0 and 0.0 values will not cause any issue.</p>\n<p>If found relevant, please upvote.</p>",
      "rawMarkdown": "I don't know much, but surely can tell that assigning 1.0 and 0.0 values will not cause any issue.\n\nIf found relevant, please upvote.",
      "votes": null
    },
    {
      "id": "2267955",
      "postDate": "05/21/2023 11:00:31",
      "content": "<p>Thanks for you answer !</p>",
      "rawMarkdown": "Thanks for you answer !",
      "votes": null
    },
    {
      "id": "2268271",
      "postDate": "05/21/2023 15:49:40",
      "content": "<p>Hi, Fares -<br>\nWe strongly recommend copying the example submission notebook and modifying it to include your own model outputs. It's difficult to get the submission format right, and difficult to debug, which is why we provide the example notebook.</p>\n<p><a href=\"https://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models\" target=\"_blank\">https://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models</a></p>",
      "rawMarkdown": "Hi, Fares -\nWe strongly recommend copying the example submission notebook and modifying it to include your own model outputs. It's difficult to get the submission format right, and difficult to debug, which is why we provide the example notebook.\n\nhttps://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models",
      "votes": null
    },
    {
      "id": "2268366",
      "postDate": "05/21/2023 17:23:21",
      "content": "<p>Hi Tom Thanks for your answer! I did that and the dataframe shape is 3x256 for the entire test file (soundscape_29201) which is not normal right ?</p>",
      "rawMarkdown": "Hi Tom Thanks for your answer! I did that and the dataframe shape is 3x256 for the entire test file (soundscape_29201) which is not normal right ?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2267945,
      "author_name": "brpuneet898",
      "author_url": "",
      "post_date": "05/21/2023 10:53:26",
      "content": "<p>I don't know much, but surely can tell that assigning 1.0 and 0.0 values will not cause any issue.</p>\n<p>If found relevant, please upvote.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2267955,
          "author_name": "faresernez",
          "author_url": "",
          "post_date": "05/21/2023 11:00:31",
          "content": "<p>Thanks for you answer !</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 2268271,
      "author_name": "tomdenton",
      "author_url": "",
      "post_date": "05/21/2023 15:49:40",
      "content": "<p>Hi, Fares -<br>\nWe strongly recommend copying the example submission notebook and modifying it to include your own model outputs. It's difficult to get the submission format right, and difficult to debug, which is why we provide the example notebook.</p>\n<p><a href=\"https://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models\" target=\"_blank\">https://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 2268366,
          "author_name": "faresernez",
          "author_url": "",
          "post_date": "05/21/2023 17:23:21",
          "content": "<p>Hi Tom Thanks for your answer! I did that and the dataframe shape is 3x256 for the entire test file (soundscape_29201) which is not normal right ?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2267922": "Hello everybody,\n\nI have been receiving a submission score error and I can't figure out what is causing it.. I am following more or less the same steps as the \"Inferring Birds with Kaggle Models\" notebook. My code outputs a 60x265 dataset of the same value types for a single file (10mn) and a 120x265 if it takes two files so it should be robust to the files replacement when scoring. I checked the value types \"np.float32\" and index = false when exporting the dataset. Does anybody have an idea what my problem could be ?\nTwo things I am doing differently are that I assign 1.0 and 0.0 to some values but I don't see why it could cause a problem and I only process 70% of the test files with my model because it is a bit slow anf for the remaining 30% I assign 4.798188e-08 to all the bird species to have a (191x60) x 265 dataset. I would be very happy if somebody lent me hand.. Also does somebody know if it possible to contact kaggle support or the organizers for this kind of issue ? Thanks a lot and have a good day :)\n\nThis is what I get if I process the given file \"soundscape_29201\" 191 times if it can help",
    "2267945": "I don't know much, but surely can tell that assigning 1.0 and 0.0 values will not cause any issue.\n\nIf found relevant, please upvote.",
    "2267955": "Thanks for you answer !",
    "2268271": "Hi, Fares -\nWe strongly recommend copying the example submission notebook and modifying it to include your own model outputs. It's difficult to get the submission format right, and difficult to debug, which is why we provide the example notebook.\n\nhttps://www.kaggle.com/code/philculliton/inferring-birds-with-kaggle-models",
    "2268366": "Hi Tom Thanks for your answer! I did that and the dataframe shape is 3x256 for the entire test file (soundscape_29201) which is not normal right ?"
  },
  "source": "meta"
}