{
  "id": 571444,
  "title": "Right now, I'm considering using Transfer Learning, as someone suggested. What else can I try? If anyone has any suggestions, please share!",
  "url": "/competitions/byu-locating-bacterial-flagellar-motors-2025/discussion/571444",
  "author_name": "",
  "post_date": "2025-04-03T09:39:32.192641400Z",
  "votes": 8,
  "comment_count": 2,
  "views": 0,
  "content": "<h4>I'm new to Kaggle competitions. I thought of this approach, but the data is taking a lot of time to train. If anyone has any ideas on how to speed it up, please let me know.</h4>\n<p><a href=\"https://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing\" target=\"_blank\">https://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing</a></p>",
  "messages": [
    {
      "id": "3169284",
      "postDate": "04/03/2025 09:39:32",
      "content": "<h4>I'm new to Kaggle competitions. I thought of this approach, but the data is taking a lot of time to train. If anyone has any ideas on how to speed it up, please let me know.</h4>\n<p><a href=\"https://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing\" target=\"_blank\">https://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing</a></p>",
      "rawMarkdown": "#### I'm new to Kaggle competitions. I thought of this approach, but the data is taking a lot of time to train. If anyone has any ideas on how to speed it up, please let me know.\n\n\nhttps://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing",
      "votes": null
    },
    {
      "id": "3175154",
      "postDate": "04/09/2025 19:27:07",
      "content": "<p>Data augmentations, data preprocessing techniques, post processing techniques, threshold tuning, varying loss functions, class weights, TTA, model ensembling, different architectures, and so on!  I hope some of these help :)</p>",
      "rawMarkdown": "Data augmentations, data preprocessing techniques, post processing techniques, threshold tuning, varying loss functions, class weights, TTA, model ensembling, different architectures, and so on!  I hope some of these help :)",
      "votes": null
    },
    {
      "id": "3181115",
      "postDate": "04/17/2025 13:45:04",
      "content": "<p>Thanks a lot! These are helpful, especially looking into TTA, class weights, and ensembling now. Appreciate the tips  !!!! </p>",
      "rawMarkdown": "Thanks a lot! These are helpful, especially looking into TTA, class weights, and ensembling now. Appreciate the tips  !!!!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3175154,
      "author_name": "connorjd",
      "author_url": "",
      "post_date": "04/09/2025 19:27:07",
      "content": "<p>Data augmentations, data preprocessing techniques, post processing techniques, threshold tuning, varying loss functions, class weights, TTA, model ensembling, different architectures, and so on!  I hope some of these help :)</p>",
      "votes": null,
      "replies": [
        {
          "id": 3181115,
          "author_name": "prothomeshmistry",
          "author_url": "",
          "post_date": "04/17/2025 13:45:04",
          "content": "<p>Thanks a lot! These are helpful, especially looking into TTA, class weights, and ensembling now. Appreciate the tips  !!!! </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3169284": "#### I'm new to Kaggle competitions. I thought of this approach, but the data is taking a lot of time to train. If anyone has any ideas on how to speed it up, please let me know.\n\n\nhttps://drive.google.com/drive/folders/1h1dn8qALaa5FjNHgB-HrMMcg8ZFK7fc4?usp=sharing",
    "3175154": "Data augmentations, data preprocessing techniques, post processing techniques, threshold tuning, varying loss functions, class weights, TTA, model ensembling, different architectures, and so on!  I hope some of these help :)",
    "3181115": "Thanks a lot! These are helpful, especially looking into TTA, class weights, and ensembling now. Appreciate the tips  !!!!"
  },
  "source": "meta"
}