{
  "id": 545820,
  "title": "Traning model on Kaggle",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/545820",
  "author_name": "",
  "post_date": "2024-11-12T10:14:43.908407600Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Have you been able to train the model on kaggle or are you doing it on your own machine? I'm having some issues here but I don't know if my machine can handle the training</p>",
  "messages": [
    {
      "id": "3043315",
      "postDate": "11/12/2024 10:14:43",
      "content": "<p>Have you been able to train the model on kaggle or are you doing it on your own machine? I'm having some issues here but I don't know if my machine can handle the training</p>",
      "rawMarkdown": "Have you been able to train the model on kaggle or are you doing it on your own machine? I'm having some issues here but I don't know if my machine can handle the training",
      "votes": null
    },
    {
      "id": "3043351",
      "postDate": "11/12/2024 11:00:14",
      "content": "<p><a href=\"https://www.kaggle.com/nickbernardini\" target=\"_blank\">@nickbernardini</a> hello,<br>\nI use Kaggle kernels only for submission of code results to my competitions and release public kernels otherwise. I train models on my local machine/ cloud providers. </p>\n<p>I suggest you can try <a href=\"runpod.io\" target=\"_blank\">runpod.io</a> for training. </p>",
      "rawMarkdown": "nickbernardini hello,\nI use Kaggle kernels only for submission of code results to my competitions and release public kernels otherwise. I train models on my local machine/ cloud providers. \n\nI suggest you can try [runpod.io](runpod.io) for training.",
      "votes": null
    },
    {
      "id": "3043503",
      "postDate": "11/12/2024 12:57:01",
      "content": "<p>30gb is not enough RAM on kaggle if you want to train on the full data. On local machine if you have 32gb+ of RAM you could give it a try because the local machine can store the exceed of the RAM into virtual memory, but there are limits too…</p>",
      "rawMarkdown": "30gb is not enough RAM on kaggle if you want to train on the full data. On local machine if you have 32gb+ of RAM you could give it a try because the local machine can store the exceed of the RAM into virtual memory, but there are limits too...",
      "votes": null
    },
    {
      "id": "3044037",
      "postDate": "11/12/2024 23:44:10",
      "content": "<p>Train MLP on Kaggle was fine.</p>",
      "rawMarkdown": "Train MLP on Kaggle was fine.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3043351,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "11/12/2024 11:00:14",
      "content": "<p><a href=\"https://www.kaggle.com/nickbernardini\" target=\"_blank\">@nickbernardini</a> hello,<br>\nI use Kaggle kernels only for submission of code results to my competitions and release public kernels otherwise. I train models on my local machine/ cloud providers. </p>\n<p>I suggest you can try <a href=\"runpod.io\" target=\"_blank\">runpod.io</a> for training. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3043503,
      "author_name": "eu1234",
      "author_url": "",
      "post_date": "11/12/2024 12:57:01",
      "content": "<p>30gb is not enough RAM on kaggle if you want to train on the full data. On local machine if you have 32gb+ of RAM you could give it a try because the local machine can store the exceed of the RAM into virtual memory, but there are limits too…</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3044037,
      "author_name": "shiyili",
      "author_url": "",
      "post_date": "11/12/2024 23:44:10",
      "content": "<p>Train MLP on Kaggle was fine.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3043315": "Have you been able to train the model on kaggle or are you doing it on your own machine? I'm having some issues here but I don't know if my machine can handle the training",
    "3043351": "nickbernardini hello,\nI use Kaggle kernels only for submission of code results to my competitions and release public kernels otherwise. I train models on my local machine/ cloud providers. \n\nI suggest you can try [runpod.io](runpod.io) for training.",
    "3043503": "30gb is not enough RAM on kaggle if you want to train on the full data. On local machine if you have 32gb+ of RAM you could give it a try because the local machine can store the exceed of the RAM into virtual memory, but there are limits too...",
    "3044037": "Train MLP on Kaggle was fine."
  },
  "source": "meta"
}