{
  "id": 461513,
  "title": "Public score 0.16",
  "url": "/competitions/UBC-OCEAN/discussion/461513",
  "author_name": "Sai Mounica Gudimella",
  "post_date": "2023-12-14T21:43:47.522000",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>I am new to submitting to competition. I only see one row in the test file and irrespective of my output label, my score is always 0.16. </p>\n<p>Can someone help how a different output is still generating 0.16 score?</p>",
  "messages": [
    {
      "id": 2561850,
      "postDate": "2023-12-14T21:43:47.523Z",
      "content": "<p>Hi,</p>\n<p>I am new to submitting to competition. I only see one row in the test file and irrespective of my output label, my score is always 0.16. </p>\n<p>Can someone help how a different output is still generating 0.16 score?</p>",
      "rawMarkdown": "Hi,\n\nI am new to submitting to competition. I only see one row in the test file and irrespective of my output label, my score is always 0.16. \n\nCan someone help how a different output is still generating 0.16 score?",
      "votes": 3
    },
    {
      "id": 2584025,
      "postDate": "2024-01-02T16:31:15.820Z",
      "content": "<p>The test file that you see is only one of the many test images.  It is there just to act as a reference but when you submit it will provide the entire test set.  There are a lot of unseen samples but the test set is larger than 1 smaple.</p>",
      "rawMarkdown": "The test file that you see is only one of the many test images.  It is there just to act as a reference but when you submit it will provide the entire test set.  There are a lot of unseen samples but the test set is larger than 1 smaple."
    },
    {
      "id": 2566031,
      "postDate": "2023-12-18T12:15:22.417Z",
      "content": "<p>I got a 0.16 score when I accidentally sent in all predictions as \"Other\" so I think it means your model is as missing it's predictions.</p>",
      "rawMarkdown": "I got a 0.16 score when I accidentally sent in all predictions as \"Other\" so I think it means your model is as missing it's predictions.",
      "replies": [
        {
          "id": 2566545,
          "postDate": "2023-12-19T00:24:54.207Z",
          "content": "<p>After receiving several \"Submission Scoring Errors\", I made some submissions to try to debug things. One of those was randomly setting the predictions. All cases scored 0.16 so I'm not sure what's going on with scoring submissions. I do suspect that my model isn't scoring very high because the training data that I'm using isn't very good.</p>",
          "rawMarkdown": "After receiving several \"Submission Scoring Errors\", I made some submissions to try to debug things. One of those was randomly setting the predictions. All cases scored 0.16 so I'm not sure what's going on with scoring submissions. I do suspect that my model isn't scoring very high because the training data that I'm using isn't very good."
        }
      ]
    },
    {
      "id": 2562639,
      "postDate": "2023-12-15T15:23:51.693Z",
      "content": "<p>I'm having a similar problem, I put up a discussion too, in my case though to test out why i would only get 0.16, i trained on a smaller part of the training data and tested the predictions on the different portion that was not taken into training, and my model would only predict the label 'CC' is what i found out, you could try the same thing and let me know</p>",
      "rawMarkdown": "I'm having a similar problem, I put up a discussion too, in my case though to test out why i would only get 0.16, i trained on a smaller part of the training data and tested the predictions on the different portion that was not taken into training, and my model would only predict the label 'CC' is what i found out, you could try the same thing and let me know"
    }
  ],
  "comments": [
    {
      "id": 2584025,
      "author_name": "Connor",
      "author_url": "",
      "post_date": "2024-01-02T16:31:15.820000",
      "content": "<p>The test file that you see is only one of the many test images.  It is there just to act as a reference but when you submit it will provide the entire test set.  There are a lot of unseen samples but the test set is larger than 1 smaple.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2566031,
      "author_name": "lmadsen",
      "author_url": "",
      "post_date": "2023-12-18T12:15:22.417000",
      "content": "<p>I got a 0.16 score when I accidentally sent in all predictions as \"Other\" so I think it means your model is as missing it's predictions.</p>",
      "votes": 0,
      "replies": [
        {
          "id": 2566545,
          "author_name": "Russ Tokuyama",
          "author_url": "",
          "post_date": "2023-12-19T00:24:54.207000",
          "content": "<p>After receiving several \"Submission Scoring Errors\", I made some submissions to try to debug things. One of those was randomly setting the predictions. All cases scored 0.16 so I'm not sure what's going on with scoring submissions. I do suspect that my model isn't scoring very high because the training data that I'm using isn't very good.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 2562639,
      "author_name": "Varun Manoj Gupta",
      "author_url": "",
      "post_date": "2023-12-15T15:23:51.693000",
      "content": "<p>I'm having a similar problem, I put up a discussion too, in my case though to test out why i would only get 0.16, i trained on a smaller part of the training data and tested the predictions on the different portion that was not taken into training, and my model would only predict the label 'CC' is what i found out, you could try the same thing and let me know</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2561850": "Hi,\n\nI am new to submitting to competition. I only see one row in the test file and irrespective of my output label, my score is always 0.16. \n\nCan someone help how a different output is still generating 0.16 score?",
    "2584025": "The test file that you see is only one of the many test images.  It is there just to act as a reference but when you submit it will provide the entire test set.  There are a lot of unseen samples but the test set is larger than 1 smaple.",
    "2566031": "I got a 0.16 score when I accidentally sent in all predictions as \"Other\" so I think it means your model is as missing it's predictions.",
    "2562639": "I'm having a similar problem, I put up a discussion too, in my case though to test out why i would only get 0.16, i trained on a smaller part of the training data and tested the predictions on the different portion that was not taken into training, and my model would only predict the label 'CC' is what i found out, you could try the same thing and let me know"
  }
}