{
  "id": 497311,
  "title": "About submit ",
  "url": "/competitions/birdclef-2024/discussion/497311",
  "author_name": "",
  "post_date": "2024-04-24T08:21:43.046605Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I want to confirm that an audio must be inferred five times? This will increase the inference speed to 5 times. My submission will time out. Is there a better way to help me?</p>",
  "messages": [
    {
      "id": "2771398",
      "postDate": "04/24/2024 08:21:43",
      "content": "<p>I want to confirm that an audio must be inferred five times? This will increase the inference speed to 5 times. My submission will time out. Is there a better way to help me?</p>",
      "rawMarkdown": "I want to confirm that an audio must be inferred five times? This will increase the inference speed to 5 times. My submission will time out. Is there a better way to help me?",
      "votes": null
    },
    {
      "id": "2771627",
      "postDate": "04/24/2024 10:37:02",
      "content": "<p>You need to score every 5 seconds slice of every test soundscape. How you score is open.</p>\n<p>Maybe you are confused because some notebooks use 5 folds models. It means that. the model was trained 5 times on slightly different data. Then they use the 5 models to score each test data slices.</p>\n<p>But you can very well use only one model.</p>",
      "rawMarkdown": "You need to score every 5 seconds slice of every test soundscape. How you score is open.\n\nMaybe you are confused because some notebooks use 5 folds models. It means that. the model was trained 5 times on slightly different data. Then they use the 5 models to score each test data slices.\n\nBut you can very well use only one model.",
      "votes": null
    },
    {
      "id": "2773987",
      "postDate": "04/25/2024 02:34:56",
      "content": "<p>Thanks for your reply！ In other words, if I have a model, when I submit it, I will only give one result for each audio?</p>",
      "rawMarkdown": "Thanks for your reply！ In other words, if I have a model, when I submit it, I will only give one result for each audio?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2771627,
      "author_name": "cpmpml",
      "author_url": "",
      "post_date": "04/24/2024 10:37:02",
      "content": "<p>You need to score every 5 seconds slice of every test soundscape. How you score is open.</p>\n<p>Maybe you are confused because some notebooks use 5 folds models. It means that. the model was trained 5 times on slightly different data. Then they use the 5 models to score each test data slices.</p>\n<p>But you can very well use only one model.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2773987,
          "author_name": "deepshark",
          "author_url": "",
          "post_date": "04/25/2024 02:34:56",
          "content": "<p>Thanks for your reply！ In other words, if I have a model, when I submit it, I will only give one result for each audio?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2771398": "I want to confirm that an audio must be inferred five times? This will increase the inference speed to 5 times. My submission will time out. Is there a better way to help me?",
    "2771627": "You need to score every 5 seconds slice of every test soundscape. How you score is open.\n\nMaybe you are confused because some notebooks use 5 folds models. It means that. the model was trained 5 times on slightly different data. Then they use the 5 models to score each test data slices.\n\nBut you can very well use only one model.",
    "2773987": "Thanks for your reply！ In other words, if I have a model, when I submit it, I will only give one result for each audio?"
  },
  "source": "meta"
}