{
  "id": 270607,
  "title": "Help needed for finding the reason for \"Notebook Threw Exception\" in my code attached below",
  "url": "/competitions/landmark-recognition-2021/discussion/270607",
  "author_name": "",
  "post_date": "2021-09-06T09:45:22.854290900Z",
  "votes": 3,
  "comment_count": 7,
  "views": 0,
  "content": "<p>I have no idea why I run into the exception: \"Notebook Threw Exception\".<br>\nMy code is attached below:<br>\n<a href=\"https://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572\" target=\"_blank\">https://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572</a><br>\nCould you please help me find the reason for this exception?</p>\n<p>Thank you very much in advance!</p>",
  "messages": [
    {
      "id": "1504344",
      "postDate": "09/06/2021 09:45:22",
      "content": "<p>I have no idea why I run into the exception: \"Notebook Threw Exception\".<br>\nMy code is attached below:<br>\n<a href=\"https://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572\" target=\"_blank\">https://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572</a><br>\nCould you please help me find the reason for this exception?</p>\n<p>Thank you very much in advance!</p>",
      "rawMarkdown": "I have no idea why I run into the exception: \"Notebook Threw Exception\".\nMy code is attached below:\nhttps://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572\nCould you please help me find the reason for this exception?\n\nThank you very much in advance!",
      "votes": null
    },
    {
      "id": "1504646",
      "postDate": "09/06/2021 14:35:25",
      "content": "<p>The private training set is different from the public one. Take care of not training+inference on the same model. I recommendo you to train the model in one notebook, download whe weights and perform the inference in another notebook. Good luck!</p>",
      "rawMarkdown": "The private training set is different from the public one. Take care of not training+inference on the same model. I recommendo you to train the model in one notebook, download whe weights and perform the inference in another notebook. Good luck!",
      "votes": null
    },
    {
      "id": "1505426",
      "postDate": "09/07/2021 08:51:00",
      "content": "<p>But there might be labels that are in the hidden test set that are not inside the public training set.<br>\nIsn't our time should be invested into finding ways for solving the problem?<br>\nIt seems like in this competition most of our time is invested in order to play a role of a detective, it is unfortunate that this is the feeling one gets.</p>",
      "rawMarkdown": "But there might be labels that are in the hidden test set that are not inside the public training set.\nIsn't our time should be invested into finding ways for solving the problem?\nIt seems like in this competition most of our time is invested in order to play a role of a detective, it is unfortunate that this is the feeling one gets.",
      "votes": null
    },
    {
      "id": "1505460",
      "postDate": "09/07/2021 09:23:58",
      "content": "<p>Hi I answered the similar question of yours in this comment. The point is that it is better not to use <code>train.csv</code> in your inference notebook. Hope this will help. 😊<br>\n<a href=\"https://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754\" target=\"_blank\">https://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754</a></p>",
      "rawMarkdown": "Hi I answered the similar question of yours in this comment. The point is that it is better not to use `train.csv` in your inference notebook. Hope this will help. 😊\nhttps://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754",
      "votes": null
    },
    {
      "id": "1505479",
      "postDate": "09/07/2021 09:48:38",
      "content": "<p>This way I will not be able to use the hidden training set for helping my model and will only rely on the public training set for the model training.<br>\nAs without using the train.csv I am not able to tell the model what the labels of the hidden training images are while training.</p>",
      "rawMarkdown": "This way I will not be able to use the hidden training set for helping my model and will only rely on the public training set for the model training.\nAs without using the train.csv I am not able to tell the model what the labels of the hidden training images are while training.",
      "votes": null
    },
    {
      "id": "1505544",
      "postDate": "09/07/2021 10:52:06",
      "content": "<p>I have tried your suggestion, could you take a look?<br>\n<a href=\"https://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402\" target=\"_blank\">https://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402</a><br>\nSomehow still the same exception occurs</p>",
      "rawMarkdown": "I have tried your suggestion, could you take a look?\nhttps://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402\nSomehow still the same exception occurs",
      "votes": null
    },
    {
      "id": "1509220",
      "postDate": "09/11/2021 03:18:51",
      "content": "<p>Hello! I cannot see the link you provided. I think the notebook status is set to private.</p>",
      "rawMarkdown": "Hello! I cannot see the link you provided. I think the notebook status is set to private.",
      "votes": null
    },
    {
      "id": "1509565",
      "postDate": "09/11/2021 12:19:39",
      "content": "<p>Hello!</p>\n<p>I've changed it, is it okay now?</p>",
      "rawMarkdown": "Hello!\n\nI've changed it, is it okay now?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1504646,
      "author_name": "enric1296",
      "author_url": "",
      "post_date": "09/06/2021 14:35:25",
      "content": "<p>The private training set is different from the public one. Take care of not training+inference on the same model. I recommendo you to train the model in one notebook, download whe weights and perform the inference in another notebook. Good luck!</p>",
      "votes": null,
      "replies": [
        {
          "id": 1505426,
          "author_name": "nizado",
          "author_url": "",
          "post_date": "09/07/2021 08:51:00",
          "content": "<p>But there might be labels that are in the hidden test set that are not inside the public training set.<br>\nIsn't our time should be invested into finding ways for solving the problem?<br>\nIt seems like in this competition most of our time is invested in order to play a role of a detective, it is unfortunate that this is the feeling one gets.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1505460,
      "author_name": "hdsk38",
      "author_url": "",
      "post_date": "09/07/2021 09:23:58",
      "content": "<p>Hi I answered the similar question of yours in this comment. The point is that it is better not to use <code>train.csv</code> in your inference notebook. Hope this will help. 😊<br>\n<a href=\"https://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754\" target=\"_blank\">https://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 1505479,
          "author_name": "nizado",
          "author_url": "",
          "post_date": "09/07/2021 09:48:38",
          "content": "<p>This way I will not be able to use the hidden training set for helping my model and will only rely on the public training set for the model training.<br>\nAs without using the train.csv I am not able to tell the model what the labels of the hidden training images are while training.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1505544,
          "author_name": "nizado",
          "author_url": "",
          "post_date": "09/07/2021 10:52:06",
          "content": "<p>I have tried your suggestion, could you take a look?<br>\n<a href=\"https://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402\" target=\"_blank\">https://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402</a><br>\nSomehow still the same exception occurs</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1509220,
          "author_name": "hdsk38",
          "author_url": "",
          "post_date": "09/11/2021 03:18:51",
          "content": "<p>Hello! I cannot see the link you provided. I think the notebook status is set to private.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1509565,
          "author_name": "nizado",
          "author_url": "",
          "post_date": "09/11/2021 12:19:39",
          "content": "<p>Hello!</p>\n<p>I've changed it, is it okay now?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1504344": "I have no idea why I run into the exception: \"Notebook Threw Exception\".\nMy code is attached below:\nhttps://www.kaggle.com/nizado/google-landmark-recognition?scriptVersionId=74101572\nCould you please help me find the reason for this exception?\n\nThank you very much in advance!",
    "1504646": "The private training set is different from the public one. Take care of not training+inference on the same model. I recommendo you to train the model in one notebook, download whe weights and perform the inference in another notebook. Good luck!",
    "1505426": "But there might be labels that are in the hidden test set that are not inside the public training set.\nIsn't our time should be invested into finding ways for solving the problem?\nIt seems like in this competition most of our time is invested in order to play a role of a detective, it is unfortunate that this is the feeling one gets.",
    "1505460": "Hi I answered the similar question of yours in this comment. The point is that it is better not to use `train.csv` in your inference notebook. Hope this will help. 😊\nhttps://www.kaggle.com/hdsk38/pytorch-starter-inference-efficientnet#1493754",
    "1505479": "This way I will not be able to use the hidden training set for helping my model and will only rely on the public training set for the model training.\nAs without using the train.csv I am not able to tell the model what the labels of the hidden training images are while training.",
    "1505544": "I have tried your suggestion, could you take a look?\nhttps://www.kaggle.com/nizado/test-submission/data?scriptVersionId=74181402\nSomehow still the same exception occurs",
    "1509220": "Hello! I cannot see the link you provided. I think the notebook status is set to private.",
    "1509565": "Hello!\n\nI've changed it, is it okay now?"
  },
  "source": "meta"
}