{
  "id": 598973,
  "title": "Request for Server Specs and Submission Error Logs",
  "url": "/competitions/rsna-intracranial-aneurysm-detection/discussion/598973",
  "author_name": "Wounsuk Rhee, M.D.",
  "post_date": "2025-08-13T13:21:27.101000",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>I was wondering if you could share the computational specifications of the remote inference server (e.g., GPU model, number of CPU cores, RAM). This would help us better estimate inference time and optimize our models accordingly before making submissions.</p>\n<p>Additionally, for submissions that result in an error or exception, would it be possible to access the related logs or error messages? Having more detailed feedback on what went wrong would greatly help in diagnosing and fixing the issue before resubmitting.</p>\n<p>Thanks a lot for your support and for hosting this competition!</p>\n<p>Wounsuk</p>",
  "messages": [
    {
      "id": 3270068,
      "postDate": "2025-08-15T16:08:34.113Z",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/wounsukrhee\" target=\"_blank\">@wounsukrhee</a>,</p>\n<p>The technical specifications for our machines are listed in the documentation here: <a href=\"https://www.kaggle.com/docs/notebooks#technical-specifications\" target=\"_blank\">Technical Specifications</a>. Inference takes place on the same kind of machine that you use to edit the submission script. For instance, if you select a T4x2 machine in the notebook editor, then a T4x2 machine will run your submission script.</p>\n<p>We unfortunately aren't able to release detailed submission logs as they can be used to extract information from the test set.</p>",
      "rawMarkdown": "Hi @wounsukrhee,\n\nThe technical specifications for our machines are listed in the documentation here: [Technical Specifications](https://www.kaggle.com/docs/notebooks#technical-specifications). Inference takes place on the same kind of machine that you use to edit the submission script. For instance, if you select a T4x2 machine in the notebook editor, then a T4x2 machine will run your submission script.\n\nWe unfortunately aren't able to release detailed submission logs as they can be used to extract information from the test set.",
      "votes": 3,
      "replies": [
        {
          "id": 3270194,
          "postDate": "2025-08-15T23:42:56.230Z",
          "content": "<p>Thank you for the detailed explanation!</p>",
          "rawMarkdown": "Thank you for the detailed explanation!"
        }
      ]
    },
    {
      "id": 3268862,
      "postDate": "2025-08-13T14:27:07.530Z",
      "content": "<p>Computational specifications should be the same as that of the one which was used to run the sample submission. That is, CPU, T4 GPU or P100 GPU.</p>",
      "rawMarkdown": "Computational specifications should be the same as that of the one which was used to run the sample submission. That is, CPU, T4 GPU or P100 GPU.",
      "votes": 1,
      "replies": [
        {
          "id": 3268923,
          "postDate": "2025-08-13T15:39:00.560Z",
          "content": "<p>Thanks for your advice</p>",
          "rawMarkdown": "Thanks for your advice"
        }
      ]
    },
    {
      "id": 3268836,
      "postDate": "2025-08-13T13:21:27.103Z",
      "content": "<p>Hi,</p>\n<p>I was wondering if you could share the computational specifications of the remote inference server (e.g., GPU model, number of CPU cores, RAM). This would help us better estimate inference time and optimize our models accordingly before making submissions.</p>\n<p>Additionally, for submissions that result in an error or exception, would it be possible to access the related logs or error messages? Having more detailed feedback on what went wrong would greatly help in diagnosing and fixing the issue before resubmitting.</p>\n<p>Thanks a lot for your support and for hosting this competition!</p>\n<p>Wounsuk</p>",
      "rawMarkdown": "Hi,\n\nI was wondering if you could share the computational specifications of the remote inference server (e.g., GPU model, number of CPU cores, RAM). This would help us better estimate inference time and optimize our models accordingly before making submissions.\n\nAdditionally, for submissions that result in an error or exception, would it be possible to access the related logs or error messages? Having more detailed feedback on what went wrong would greatly help in diagnosing and fixing the issue before resubmitting.\n\nThanks a lot for your support and for hosting this competition!\n\nWounsuk",
      "votes": 2
    }
  ],
  "comments": [
    {
      "id": 3270068,
      "author_name": "Ryan Holbrook",
      "author_url": "",
      "post_date": "2025-08-15T16:08:34.113000",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/wounsukrhee\" target=\"_blank\">@wounsukrhee</a>,</p>\n<p>The technical specifications for our machines are listed in the documentation here: <a href=\"https://www.kaggle.com/docs/notebooks#technical-specifications\" target=\"_blank\">Technical Specifications</a>. Inference takes place on the same kind of machine that you use to edit the submission script. For instance, if you select a T4x2 machine in the notebook editor, then a T4x2 machine will run your submission script.</p>\n<p>We unfortunately aren't able to release detailed submission logs as they can be used to extract information from the test set.</p>",
      "votes": 3,
      "replies": [
        {
          "id": 3270194,
          "author_name": "Wounsuk Rhee, M.D.",
          "author_url": "",
          "post_date": "2025-08-15T23:42:56.230000",
          "content": "<p>Thank you for the detailed explanation!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3268862,
      "author_name": "coderRKJ",
      "author_url": "",
      "post_date": "2025-08-13T14:27:07.530000",
      "content": "<p>Computational specifications should be the same as that of the one which was used to run the sample submission. That is, CPU, T4 GPU or P100 GPU.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 3268923,
          "author_name": "Wounsuk Rhee, M.D.",
          "author_url": "",
          "post_date": "2025-08-13T15:39:00.560000",
          "content": "<p>Thanks for your advice</p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3270068": "Hi @wounsukrhee,\n\nThe technical specifications for our machines are listed in the documentation here: [Technical Specifications](https://www.kaggle.com/docs/notebooks#technical-specifications). Inference takes place on the same kind of machine that you use to edit the submission script. For instance, if you select a T4x2 machine in the notebook editor, then a T4x2 machine will run your submission script.\n\nWe unfortunately aren't able to release detailed submission logs as they can be used to extract information from the test set.",
    "3268862": "Computational specifications should be the same as that of the one which was used to run the sample submission. That is, CPU, T4 GPU or P100 GPU.",
    "3268836": "Hi,\n\nI was wondering if you could share the computational specifications of the remote inference server (e.g., GPU model, number of CPU cores, RAM). This would help us better estimate inference time and optimize our models accordingly before making submissions.\n\nAdditionally, for submissions that result in an error or exception, would it be possible to access the related logs or error messages? Having more detailed feedback on what went wrong would greatly help in diagnosing and fixing the issue before resubmitting.\n\nThanks a lot for your support and for hosting this competition!\n\nWounsuk"
  }
}