{
  "id": 87634,
  "title": "Window vs batch size ",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/87634",
  "author_name": "carolcuesta",
  "post_date": "2019-04-02T07:05:41.929000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi all!</p>\n\n<p>For those working on deep learning models, what are your best model's window and batch sizes? We have found them to be very important, so far we use windows 15000 data points and a batch size of 64, and some pooling on the first layer. </p>\n\n<p>Thank you in advance </p>",
  "messages": [
    {
      "id": 505540,
      "postDate": "2019-04-02T07:05:41.930Z",
      "content": "<p>Hi all!</p>\n\n<p>For those working on deep learning models, what are your best model's window and batch sizes? We have found them to be very important, so far we use windows 15000 data points and a batch size of 64, and some pooling on the first layer. </p>\n\n<p>Thank you in advance </p>",
      "rawMarkdown": "Hi all!\n\nFor those working on deep learning models, what are your best model's window and batch sizes? We have found them to be very important, so far we use windows 15000 data points and a batch size of 64, and some pooling on the first layer. \n\nThank you in advance ",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "505540": "Hi all!\n\nFor those working on deep learning models, what are your best model's window and batch sizes? We have found them to be very important, so far we use windows 15000 data points and a batch size of 64, and some pooling on the first layer. \n\nThank you in advance "
  }
}