{
  "id": 496987,
  "title": "is there any framework to build chunkwise models ",
  "url": "/competitions/leap-atmospheric-physics-ai-climsim/discussion/496987",
  "author_name": "Satej Raste",
  "post_date": "2024-04-23T06:53:58.111000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>i am doing chunkwise reading whatever i do parametric models or boosting models wont be update parameters chunk wise is there a way i can update my model after each chunk with new data as training data or is neural network is the only possible approach ? </p>",
  "messages": [
    {
      "id": 2769087,
      "postDate": "2024-04-23T06:53:58.110Z",
      "content": "<p>i am doing chunkwise reading whatever i do parametric models or boosting models wont be update parameters chunk wise is there a way i can update my model after each chunk with new data as training data or is neural network is the only possible approach ? </p>",
      "rawMarkdown": "i am doing chunkwise reading whatever i do parametric models or boosting models wont be update parameters chunk wise is there a way i can update my model after each chunk with new data as training data or is neural network is the only possible approach ? \n",
      "votes": 1
    },
    {
      "id": 2769682,
      "postDate": "2024-04-23T13:56:45.530Z",
      "content": "<p>Might be of interest: </p>\n<p>LightGBM: train() vs update() vs refit() -  <a href=\"https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit\" target=\"_blank\">https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit</a></p>",
      "rawMarkdown": "Might be of interest: \n\nLightGBM: train() vs update() vs refit() -  https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit",
      "votes": 2,
      "replies": [
        {
          "id": 2770285,
          "postDate": "2024-04-23T19:00:53.383Z",
          "content": "<p>Thank you for the link! That could be very useful, I learned something new</p>",
          "rawMarkdown": "Thank you for the link! That could be very useful, I learned something new"
        },
        {
          "id": 2770954,
          "postDate": "2024-04-24T05:03:40.603Z",
          "content": "<p>Thank you for the link its nice to know since this is parameter is available in Xgb too </p>",
          "rawMarkdown": "Thank you for the link its nice to know since this is parameter is available in Xgb too \n"
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 2769682,
      "author_name": "JM",
      "author_url": "",
      "post_date": "2024-04-23T13:56:45.530000",
      "content": "<p>Might be of interest: </p>\n<p>LightGBM: train() vs update() vs refit() -  <a href=\"https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit\" target=\"_blank\">https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit</a></p>",
      "votes": 2,
      "replies": [
        {
          "id": 2770285,
          "author_name": "menno",
          "author_url": "",
          "post_date": "2024-04-23T19:00:53.383000",
          "content": "<p>Thank you for the link! That could be very useful, I learned something new</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 2770954,
          "author_name": "Satej Raste",
          "author_url": "",
          "post_date": "2024-04-24T05:03:40.603000",
          "content": "<p>Thank you for the link its nice to know since this is parameter is available in Xgb too </p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2769087": "i am doing chunkwise reading whatever i do parametric models or boosting models wont be update parameters chunk wise is there a way i can update my model after each chunk with new data as training data or is neural network is the only possible approach ? \n",
    "2769682": "Might be of interest: \n\nLightGBM: train() vs update() vs refit() -  https://stackoverflow.com/questions/73664093/lightgbm-train-vs-update-vs-refit"
  }
}