{
  "id": 265166,
  "title": "Notebook throws unknow exception",
  "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/265166",
  "author_name": "",
  "post_date": "2021-08-14T21:27:55.509679100Z",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hello everybody, I am a beginner in Kaggle. I'm trying to use tf.data to load data from input directory and transfer learning based on efficientnet. My idea is based on Ayush Thakur's <a href=\"https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b\" target=\"_blank\">https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b</a></p>\n<p>My notebook runs successfully when I save and commit it. But everytime I submit it threw exception error. I have no idea where the bug is and where to look at. I will appreciate any suggestions. Thank you all.</p>\n<p>my notebook: <a href=\"https://www.kaggle.com/robinsooooon/braintumor-sub2\" target=\"_blank\">https://www.kaggle.com/robinsooooon/braintumor-sub2</a></p>",
  "messages": [
    {
      "id": "1472436",
      "postDate": "08/14/2021 21:27:55",
      "content": "<p>Hello everybody, I am a beginner in Kaggle. I'm trying to use tf.data to load data from input directory and transfer learning based on efficientnet. My idea is based on Ayush Thakur's <a href=\"https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b\" target=\"_blank\">https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b</a></p>\n<p>My notebook runs successfully when I save and commit it. But everytime I submit it threw exception error. I have no idea where the bug is and where to look at. I will appreciate any suggestions. Thank you all.</p>\n<p>my notebook: <a href=\"https://www.kaggle.com/robinsooooon/braintumor-sub2\" target=\"_blank\">https://www.kaggle.com/robinsooooon/braintumor-sub2</a></p>",
      "rawMarkdown": "Hello everybody, I am a beginner in Kaggle. I'm trying to use tf.data to load data from input directory and transfer learning based on efficientnet. My idea is based on Ayush Thakur's https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b\n\nMy notebook runs successfully when I save and commit it. But everytime I submit it threw exception error. I have no idea where the bug is and where to look at. I will appreciate any suggestions. Thank you all.\n\nmy notebook: https://www.kaggle.com/robinsooooon/braintumor-sub2",
      "votes": null
    },
    {
      "id": "1472489",
      "postDate": "08/14/2021 22:38:54",
      "content": "<p>Remove the \"./\" from the submission.csv file name</p>",
      "rawMarkdown": "Remove the \"./\" from the submission.csv file name",
      "votes": null
    },
    {
      "id": "1472661",
      "postDate": "08/15/2021 03:47:00",
      "content": "<p>Did you make sure you aren't relying on internet access?</p>",
      "rawMarkdown": "Did you make sure you aren't relying on internet access?",
      "votes": null
    },
    {
      "id": "1472665",
      "postDate": "08/15/2021 03:49:35",
      "content": "<p>FYI, I think you can't rely on external datasets per this discussion: <a href=\"https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612\" target=\"_blank\">https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612</a></p>\n<p>I know the rules say you can use your own data but its just weird…</p>",
      "rawMarkdown": "FYI, I think you can't rely on external datasets per this discussion: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612\n\nI know the rules say you can use your own data but its just weird...",
      "votes": null
    },
    {
      "id": "1472684",
      "postDate": "08/15/2021 04:03:10",
      "content": "<p>Emmm, since I cannot use internet but still want to use pretrained efficientnet, I have to use external dataset to load efficientnet. Maybe I should seperate training and testing into two different notebooks?</p>",
      "rawMarkdown": "Emmm, since I cannot use internet but still want to use pretrained efficientnet, I have to use external dataset to load efficientnet. Maybe I should seperate training and testing into two different notebooks?",
      "votes": null
    },
    {
      "id": "1473247",
      "postDate": "08/15/2021 13:05:25",
      "content": "<p>Hmmm what I ended up doing was just rolling my own model in PyTorch and training it. I wanted to use a pretrained model too but tbh going thru the exercise of developing my own model was very instructive.</p>",
      "rawMarkdown": "Hmmm what I ended up doing was just rolling my own model in PyTorch and training it. I wanted to use a pretrained model too but tbh going thru the exercise of developing my own model was very instructive.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1472489,
      "author_name": "richardepstein",
      "author_url": "",
      "post_date": "08/14/2021 22:38:54",
      "content": "<p>Remove the \"./\" from the submission.csv file name</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1472661,
      "author_name": "d223chen",
      "author_url": "",
      "post_date": "08/15/2021 03:47:00",
      "content": "<p>Did you make sure you aren't relying on internet access?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1472665,
      "author_name": "d223chen",
      "author_url": "",
      "post_date": "08/15/2021 03:49:35",
      "content": "<p>FYI, I think you can't rely on external datasets per this discussion: <a href=\"https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612\" target=\"_blank\">https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612</a></p>\n<p>I know the rules say you can use your own data but its just weird…</p>",
      "votes": null,
      "replies": [
        {
          "id": 1472684,
          "author_name": "robinsooooon",
          "author_url": "",
          "post_date": "08/15/2021 04:03:10",
          "content": "<p>Emmm, since I cannot use internet but still want to use pretrained efficientnet, I have to use external dataset to load efficientnet. Maybe I should seperate training and testing into two different notebooks?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1473247,
          "author_name": "d223chen",
          "author_url": "",
          "post_date": "08/15/2021 13:05:25",
          "content": "<p>Hmmm what I ended up doing was just rolling my own model in PyTorch and training it. I wanted to use a pretrained model too but tbh going thru the exercise of developing my own model was very instructive.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1472436": "Hello everybody, I am a beginner in Kaggle. I'm trying to use tf.data to load data from input directory and transfer learning based on efficientnet. My idea is based on Ayush Thakur's https://www.kaggle.com/ayuraj/train-brain-tumor-as-video-classification-w-b\n\nMy notebook runs successfully when I save and commit it. But everytime I submit it threw exception error. I have no idea where the bug is and where to look at. I will appreciate any suggestions. Thank you all.\n\nmy notebook: https://www.kaggle.com/robinsooooon/braintumor-sub2",
    "1472489": "Remove the \"./\" from the submission.csv file name",
    "1472661": "Did you make sure you aren't relying on internet access?",
    "1472665": "FYI, I think you can't rely on external datasets per this discussion: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/263612\n\nI know the rules say you can use your own data but its just weird...",
    "1472684": "Emmm, since I cannot use internet but still want to use pretrained efficientnet, I have to use external dataset to load efficientnet. Maybe I should seperate training and testing into two different notebooks?",
    "1473247": "Hmmm what I ended up doing was just rolling my own model in PyTorch and training it. I wanted to use a pretrained model too but tbh going thru the exercise of developing my own model was very instructive."
  },
  "source": "meta"
}