{
  "id": 223691,
  "title": "Issue with HPA cell segmentator - help needed!",
  "url": "/competitions/hpa-single-cell-image-classification/discussion/223691",
  "author_name": "Darek Kłeczek",
  "post_date": "2021-03-05T05:56:56.265000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I'm running into a strange issue using cell segmentator on public images, specifically image size 1728, even though I set padding to True. Can't figure out what is wrong. Would anyone be able to help? The notebook with error is here: <a href=\"https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2\" target=\"_blank\">https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2</a></p>\n<p><code>RuntimeError: The size of tensor a (9) must match the size of tensor b (3) at non-singleton dimension 1</code></p>",
  "messages": [
    {
      "id": 1227991,
      "postDate": "2021-03-06T02:25:56.537Z",
      "content": "<p>Could you solve the error?<br>\nSorry if my finding below is incorrect.</p>",
      "rawMarkdown": "Could you solve the error?\nSorry if my finding below is incorrect.",
      "votes": 1
    },
    {
      "id": 1227364,
      "postDate": "2021-03-05T13:03:41.770Z",
      "content": "<p>Hi. I took a quick look at your code and also tried some cells, and noticed 1728 images in your datasets are array([None, None, None], dtype=object).</p>",
      "rawMarkdown": "Hi. I took a quick look at your code and also tried some cells, and noticed 1728 images in your datasets are array([None, None, None], dtype=object).",
      "votes": 1,
      "replies": [
        {
          "id": 1228032,
          "postDate": "2021-03-06T03:57:14.020Z",
          "content": "<p>Thanks <a href=\"https://www.kaggle.com/drtausamaru\" target=\"_blank\">@drtausamaru</a> ! This helps a lot - the issue then needs to be in the other notebook where I create the dataset. I think I know where it is, instead of saving files to disk, I write them into a zip file and this is where probably the error happens, I should be able to troubleshoot this further. I will probably get time to work on this on Monday, when I find out I'll report here for others. I also shared the notebook where I think the issue is: <a href=\"https://www.kaggle.com/thedrcat/hpa-public-data-negative-sample/\" target=\"_blank\">https://www.kaggle.com/thedrcat/hpa-public-data-negative-sample/</a></p>",
          "rawMarkdown": "Thanks @drtausamaru ! This helps a lot - the issue then needs to be in the other notebook where I create the dataset. I think I know where it is, instead of saving files to disk, I write them into a zip file and this is where probably the error happens, I should be able to troubleshoot this further. I will probably get time to work on this on Monday, when I find out I'll report here for others. I also shared the notebook where I think the issue is: https://www.kaggle.com/thedrcat/hpa-public-data-negative-sample/",
          "votes": 1
        }
      ]
    },
    {
      "id": 1227040,
      "postDate": "2021-03-05T05:56:56.267Z",
      "content": "<p>I'm running into a strange issue using cell segmentator on public images, specifically image size 1728, even though I set padding to True. Can't figure out what is wrong. Would anyone be able to help? The notebook with error is here: <a href=\"https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2\" target=\"_blank\">https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2</a></p>\n<p><code>RuntimeError: The size of tensor a (9) must match the size of tensor b (3) at non-singleton dimension 1</code></p>",
      "rawMarkdown": "I'm running into a strange issue using cell segmentator on public images, specifically image size 1728, even though I set padding to True. Can't figure out what is wrong. Would anyone be able to help? The notebook with error is here: https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2\n\n`RuntimeError: The size of tensor a (9) must match the size of tensor b (3) at non-singleton dimension 1`",
      "votes": 1
    }
  ],
  "comments": [
    {
      "id": 1227991,
      "author_name": "cool_rabbit",
      "author_url": "",
      "post_date": "2021-03-06T02:25:56.537000",
      "content": "<p>Could you solve the error?<br>\nSorry if my finding below is incorrect.</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1227364,
      "author_name": "cool_rabbit",
      "author_url": "",
      "post_date": "2021-03-05T13:03:41.770000",
      "content": "<p>Hi. I took a quick look at your code and also tried some cells, and noticed 1728 images in your datasets are array([None, None, None], dtype=object).</p>",
      "votes": 1,
      "replies": [
        {
          "id": 1228032,
          "author_name": "Darek Kłeczek",
          "author_url": "",
          "post_date": "2021-03-06T03:57:14.020000",
          "content": "<p>Thanks <a href=\"https://www.kaggle.com/drtausamaru\" target=\"_blank\">@drtausamaru</a> ! This helps a lot - the issue then needs to be in the other notebook where I create the dataset. I think I know where it is, instead of saving files to disk, I write them into a zip file and this is where probably the error happens, I should be able to troubleshoot this further. I will probably get time to work on this on Monday, when I find out I'll report here for others. I also shared the notebook where I think the issue is: <a href=\"https://www.kaggle.com/thedrcat/hpa-public-data-negative-sample/\" target=\"_blank\">https://www.kaggle.com/thedrcat/hpa-public-data-negative-sample/</a></p>",
          "votes": 1,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1227991": "Could you solve the error?\nSorry if my finding below is incorrect.",
    "1227364": "Hi. I took a quick look at your code and also tried some cells, and noticed 1728 images in your datasets are array([None, None, None], dtype=object).",
    "1227040": "I'm running into a strange issue using cell segmentator on public images, specifically image size 1728, even though I set padding to True. Can't figure out what is wrong. Would anyone be able to help? The notebook with error is here: https://www.kaggle.com/thedrcat/hpa-negative-sample-cell-segmentation-2\n\n`RuntimeError: The size of tensor a (9) must match the size of tensor b (3) at non-singleton dimension 1`"
  }
}