{
  "id": 174435,
  "title": "Melanoma - Resource Exhausted Error ",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/174435",
  "author_name": "",
  "post_date": "2020-08-13T14:52:33.112114200Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>I had run my notebook almost a week ago with the same code snippets and it was working fine giving the required result. But now when I try to run my notebook it displaying the below results. I have tried by reducing batch_size(which I don't want because it reduces my AUC value), but still, it's taking a lot of time to even initialize the running.</p>\n<p>More Info: Notebooks are being run on TPU.</p>\n<p>ResourceExhaustedError: 9 root error(s) found.<br>\n(0) Resource exhausted: {{function_node _inferencetrainfunction575069}} Compilation failure: Ran out of memory in memory space hbm. Used 19.13G of 15.48G hbm. Exceeded hbm capacity by 3.64G.</p>\n<p>Will be glad if someones help as this the last week of competition and plus my first. So, I want to do well and get a high leaderboard score.</p>",
  "messages": [
    {
      "id": "969216",
      "postDate": "08/13/2020 14:52:33",
      "content": "<p>I had run my notebook almost a week ago with the same code snippets and it was working fine giving the required result. But now when I try to run my notebook it displaying the below results. I have tried by reducing batch_size(which I don't want because it reduces my AUC value), but still, it's taking a lot of time to even initialize the running.</p>\n<p>More Info: Notebooks are being run on TPU.</p>\n<p>ResourceExhaustedError: 9 root error(s) found.<br>\n(0) Resource exhausted: {{function_node _inferencetrainfunction575069}} Compilation failure: Ran out of memory in memory space hbm. Used 19.13G of 15.48G hbm. Exceeded hbm capacity by 3.64G.</p>\n<p>Will be glad if someones help as this the last week of competition and plus my first. So, I want to do well and get a high leaderboard score.</p>",
      "rawMarkdown": "I had run my notebook almost a week ago with the same code snippets and it was working fine giving the required result. But now when I try to run my notebook it displaying the below results. I have tried by reducing batch_size(which I don't want because it reduces my AUC value), but still, it's taking a lot of time to even initialize the running.\n\nMore Info: Notebooks are being run on TPU.\n\nResourceExhaustedError: 9 root error(s) found.\n(0) Resource exhausted: {{function_node _inferencetrainfunction575069}} Compilation failure: Ran out of memory in memory space hbm. Used 19.13G of 15.48G hbm. Exceeded hbm capacity by 3.64G.\n\nWill be glad if someones help as this the last week of competition and plus my first. So, I want to do well and get a high leaderboard score.",
      "votes": null
    },
    {
      "id": "969251",
      "postDate": "08/13/2020 15:15:17",
      "content": "<p>pytorch????</p>",
      "rawMarkdown": "pytorch????",
      "votes": null
    },
    {
      "id": "969292",
      "postDate": "08/13/2020 15:40:53",
      "content": "<p>Everyone is having this problem. Discussion <a href=\"https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/174299\" target=\"_blank\">here</a></p>",
      "rawMarkdown": "Everyone is having this problem. Discussion [here][1]\n\n[1]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/174299",
      "votes": null
    },
    {
      "id": "969336",
      "postDate": "08/13/2020 16:07:34",
      "content": "<p>No sir, I am using Tensorflow.</p>",
      "rawMarkdown": "No sir, I am using Tensorflow.",
      "votes": null
    },
    {
      "id": "969341",
      "postDate": "08/13/2020 16:09:32",
      "content": "<p>Thank you very much, sir I will go through the discussion.</p>",
      "rawMarkdown": "Thank you very much, sir I will go through the discussion.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 969251,
      "author_name": "abhishek",
      "author_url": "",
      "post_date": "08/13/2020 15:15:17",
      "content": "<p>pytorch????</p>",
      "votes": null,
      "replies": [
        {
          "id": 969336,
          "author_name": "vishnu123",
          "author_url": "",
          "post_date": "08/13/2020 16:07:34",
          "content": "<p>No sir, I am using Tensorflow.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 969292,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "08/13/2020 15:40:53",
      "content": "<p>Everyone is having this problem. Discussion <a href=\"https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/174299\" target=\"_blank\">here</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 969341,
          "author_name": "vishnu123",
          "author_url": "",
          "post_date": "08/13/2020 16:09:32",
          "content": "<p>Thank you very much, sir I will go through the discussion.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "969216": "I had run my notebook almost a week ago with the same code snippets and it was working fine giving the required result. But now when I try to run my notebook it displaying the below results. I have tried by reducing batch_size(which I don't want because it reduces my AUC value), but still, it's taking a lot of time to even initialize the running.\n\nMore Info: Notebooks are being run on TPU.\n\nResourceExhaustedError: 9 root error(s) found.\n(0) Resource exhausted: {{function_node _inferencetrainfunction575069}} Compilation failure: Ran out of memory in memory space hbm. Used 19.13G of 15.48G hbm. Exceeded hbm capacity by 3.64G.\n\nWill be glad if someones help as this the last week of competition and plus my first. So, I want to do well and get a high leaderboard score.",
    "969251": "pytorch????",
    "969292": "Everyone is having this problem. Discussion [here][1]\n\n[1]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/174299",
    "969336": "No sir, I am using Tensorflow.",
    "969341": "Thank you very much, sir I will go through the discussion."
  },
  "source": "meta"
}