{
  "id": 679053,
  "title": "Any workaround to submit inference notebook without gpu quota left?",
  "url": "/competitions/vesuvius-challenge-surface-detection/discussion/679053",
  "author_name": "",
  "post_date": "2026-02-26T19:49:24.313051600Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hope everyone is doing great,</p>\n<p>I just finished training a model, and ran out of Kaggle gpu quota. Now i want to submit the inferencing notebook but it's not possible because of no quota left. As the inference time is around 6 to 8 mins, Is there any way to submit these notebooks?</p>\n<p>(The model was based on nnUNetv2, has someone tried it's inferencing on TPU?)</p>\n<p>I don't have colab pro, so can't look at that approach too. </p>\n<p>Any valuable feedback would be appreciated. Thanks in advance.</p>",
  "messages": [
    {
      "id": "3414444",
      "postDate": "02/26/2026 19:49:24",
      "content": "<p>Hope everyone is doing great,</p>\n<p>I just finished training a model, and ran out of Kaggle gpu quota. Now i want to submit the inferencing notebook but it's not possible because of no quota left. As the inference time is around 6 to 8 mins, Is there any way to submit these notebooks?</p>\n<p>(The model was based on nnUNetv2, has someone tried it's inferencing on TPU?)</p>\n<p>I don't have colab pro, so can't look at that approach too. </p>\n<p>Any valuable feedback would be appreciated. Thanks in advance.</p>",
      "rawMarkdown": "Hope everyone is doing great,\n\nI just finished training a model, and ran out of Kaggle gpu quota. Now i want to submit the inferencing notebook but it's not possible because of no quota left. As the inference time is around 6 to 8 mins, Is there any way to submit these notebooks?\n\n(The model was based on nnUNetv2, has someone tried it's inferencing on TPU?)\n\nI don't have colab pro, so can't look at that approach too. \n\nAny valuable feedback would be appreciated. Thanks in advance.",
      "votes": null
    },
    {
      "id": "3414466",
      "postDate": "02/26/2026 20:45:44",
      "content": "<p>Try to submit the output . The TPU is not accepted : this is the mesage i got :)</p>",
      "rawMarkdown": "Try to submit the output . The TPU is not accepted : this is the mesage i got :)",
      "votes": null
    },
    {
      "id": "3414473",
      "postDate": "02/26/2026 21:06:50",
      "content": "<p>Sorry but as far I know. Code competitions need the code to start to run before trigger hidden execution. So no.</p>",
      "rawMarkdown": "Sorry but as far I know. Code competitions need the code to start to run before trigger hidden execution. So no.",
      "votes": null
    },
    {
      "id": "3414478",
      "postDate": "02/26/2026 21:36:32",
      "content": "<p>I haven't run inference notebook yet, so can't submit any output. </p>",
      "rawMarkdown": "I haven't run inference notebook yet, so can't submit any output.",
      "votes": null
    },
    {
      "id": "3414702",
      "postDate": "02/27/2026 13:27:30",
      "content": "<p>buy  Colab Pro now</p>",
      "rawMarkdown": "buy  Colab Pro now",
      "votes": null
    },
    {
      "id": "3414708",
      "postDate": "02/27/2026 13:46:08",
      "content": "<p>Colab pro is giving additional 15 hours of GPU on kaggle </p>",
      "rawMarkdown": "Colab pro is giving additional 15 hours of GPU on kaggle",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3414466,
      "author_name": "crischir",
      "author_url": "",
      "post_date": "02/26/2026 20:45:44",
      "content": "<p>Try to submit the output . The TPU is not accepted : this is the mesage i got :)</p>",
      "votes": null,
      "replies": [
        {
          "id": 3414478,
          "author_name": "muhammadibrahim3093",
          "author_url": "",
          "post_date": "02/26/2026 21:36:32",
          "content": "<p>I haven't run inference notebook yet, so can't submit any output. </p>",
          "votes": null,
          "replies": [
            {
              "id": 3414708,
              "author_name": "crischir",
              "author_url": "",
              "post_date": "02/27/2026 13:46:08",
              "content": "<p>Colab pro is giving additional 15 hours of GPU on kaggle </p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 3414473,
      "author_name": "sacuscreed",
      "author_url": "",
      "post_date": "02/26/2026 21:06:50",
      "content": "<p>Sorry but as far I know. Code competitions need the code to start to run before trigger hidden execution. So no.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3414702,
      "author_name": "aaaabbs",
      "author_url": "",
      "post_date": "02/27/2026 13:27:30",
      "content": "<p>buy  Colab Pro now</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3414444": "Hope everyone is doing great,\n\nI just finished training a model, and ran out of Kaggle gpu quota. Now i want to submit the inferencing notebook but it's not possible because of no quota left. As the inference time is around 6 to 8 mins, Is there any way to submit these notebooks?\n\n(The model was based on nnUNetv2, has someone tried it's inferencing on TPU?)\n\nI don't have colab pro, so can't look at that approach too. \n\nAny valuable feedback would be appreciated. Thanks in advance.",
    "3414466": "Try to submit the output . The TPU is not accepted : this is the mesage i got :)",
    "3414473": "Sorry but as far I know. Code competitions need the code to start to run before trigger hidden execution. So no.",
    "3414478": "I haven't run inference notebook yet, so can't submit any output.",
    "3414702": "buy  Colab Pro now",
    "3414708": "Colab pro is giving additional 15 hours of GPU on kaggle"
  },
  "source": "meta"
}