{
  "id": 612170,
  "title": "gpu requirements",
  "url": "/competitions/cafa-6-protein-function-prediction/discussion/612170",
  "author_name": "Taha_Alshatiri",
  "post_date": "2025-10-17T09:41:34.196000",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>am new to such topics and I want to ask before joining, what kind of gpu might be required in this competition</p>",
  "messages": [
    {
      "id": 3303254,
      "postDate": "2025-10-17T14:07:10.253Z",
      "content": "<p>GPU is not a must-have requirement for the competition, but depending on your prediction method you might need one. Three benchmark methods listed in the leaderboard were created without GPU.</p>",
      "rawMarkdown": "GPU is not a must-have requirement for the competition, but depending on your prediction method you might need one. Three benchmark methods listed in the leaderboard were created without GPU.",
      "votes": 4
    },
    {
      "id": 3303172,
      "postDate": "2025-10-17T09:41:34.197Z",
      "content": "<p>am new to such topics and I want to ask before joining, what kind of gpu might be required in this competition</p>",
      "rawMarkdown": "am new to such topics and I want to ask before joining, what kind of gpu might be required in this competition",
      "votes": 2
    },
    {
      "id": 3303902,
      "postDate": "2025-10-19T08:06:20.033Z",
      "content": "<p>On Kaggle, you can use their free GPU resources, but usage is limited, so manage sessions carefully. You can also use a pre-trained model as a feature extractor to save compute and speed up training.</p>",
      "rawMarkdown": "On Kaggle, you can use their free GPU resources, but usage is limited, so manage sessions carefully. You can also use a pre-trained model as a feature extractor to save compute and speed up training."
    },
    {
      "id": 3304199,
      "postDate": "2025-10-20T04:37:26.470Z",
      "content": "<p>there is couple of GPU options under settings in Kaggle Notebook. You can run it using CPU too</p>",
      "rawMarkdown": "there is couple of GPU options under settings in Kaggle Notebook. You can run it using CPU too",
      "votes": -1
    },
    {
      "id": 3303900,
      "postDate": "2025-10-19T08:03:11.220Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 3303254,
      "author_name": "An Phan",
      "author_url": "",
      "post_date": "2025-10-17T14:07:10.253000",
      "content": "<p>GPU is not a must-have requirement for the competition, but depending on your prediction method you might need one. Three benchmark methods listed in the leaderboard were created without GPU.</p>",
      "votes": 4,
      "replies": []
    },
    {
      "id": 3303902,
      "author_name": "Rishabh_Jain_Dev",
      "author_url": "",
      "post_date": "2025-10-19T08:06:20.033000",
      "content": "<p>On Kaggle, you can use their free GPU resources, but usage is limited, so manage sessions carefully. You can also use a pre-trained model as a feature extractor to save compute and speed up training.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3304199,
      "author_name": "Elmaddin Guliyev",
      "author_url": "",
      "post_date": "2025-10-20T04:37:26.470000",
      "content": "<p>there is couple of GPU options under settings in Kaggle Notebook. You can run it using CPU too</p>",
      "votes": -1,
      "replies": []
    },
    {
      "id": 3303900,
      "author_name": "",
      "author_url": "",
      "post_date": "2025-10-19T08:03:11.220000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3303254": "GPU is not a must-have requirement for the competition, but depending on your prediction method you might need one. Three benchmark methods listed in the leaderboard were created without GPU.",
    "3303172": "am new to such topics and I want to ask before joining, what kind of gpu might be required in this competition",
    "3303902": "On Kaggle, you can use their free GPU resources, but usage is limited, so manage sessions carefully. You can also use a pre-trained model as a feature extractor to save compute and speed up training.",
    "3304199": "there is couple of GPU options under settings in Kaggle Notebook. You can run it using CPU too",
    "3303900": ""
  }
}