{
  "id": 100069,
  "title": "are you having problems to add .pth files to your kernels? (using Kaggle cli)",
  "url": "/competitions/aptos2019-blindness-detection/discussion/100069",
  "author_name": "",
  "post_date": "2019-07-16T12:33:18.383936600Z",
  "votes": 3,
  "comment_count": 10,
  "views": 0,
  "content": "<p>are you experiencing problems adding .pth files to your kernels?</p>\n\n<p><a href=\"https://github.com/Kaggle/kaggle-api/issues/202\">https://github.com/Kaggle/kaggle-api/issues/202</a></p>\n\n<p>without using pretrained checkpoints, it's quite difficult to participate in image related competitions</p>",
  "messages": [
    {
      "id": "577164",
      "postDate": "07/16/2019 12:33:18",
      "content": "<p>are you experiencing problems adding .pth files to your kernels?</p>\n\n<p><a href=\"https://github.com/Kaggle/kaggle-api/issues/202\">https://github.com/Kaggle/kaggle-api/issues/202</a></p>\n\n<p>without using pretrained checkpoints, it's quite difficult to participate in image related competitions</p>",
      "rawMarkdown": "are you experiencing problems adding .pth files to your kernels?\n\nhttps://github.com/Kaggle/kaggle-api/issues/202\n\nwithout using pretrained checkpoints, it's quite difficult to participate in image related competitions",
      "votes": null
    },
    {
      "id": "577179",
      "postDate": "07/16/2019 12:53:29",
      "content": "<p>No</p>",
      "rawMarkdown": "No",
      "votes": null
    },
    {
      "id": "577190",
      "postDate": "07/16/2019 13:15:18",
      "content": "<p>Hey, can you explain specifically what problem are you facing?</p>",
      "rawMarkdown": "Hey, can you explain specifically what problem are you facing?",
      "votes": null
    },
    {
      "id": "577205",
      "postDate": "07/16/2019 13:29:18",
      "content": "<p>I added this kaggle.com/hmendonca/efficientnet-pytorch datasource to my kernel.</p>\n\n<p>It has some .pth (pytorch checkpoint) files.</p>\n\n<p>The dataset is correctly added to the kernel, but all the .pth files have a question mark symbol.\nWhen I click the question mark, I can read: .pth is not a supported format.</p>\n\n<p>The same happens with .pkl files (pickle)</p>\n\n<p>And the same happens if I try to add it as another kernel output (instead of a dataset)</p>\n\n<p>I'm using Kaggle cli to create the kernel.  And I've checked that other files with extensions like .csv are working</p>",
      "rawMarkdown": "I added this kaggle.com/hmendonca/efficientnet-pytorch datasource to my kernel.\n\nIt has some .pth (pytorch checkpoint) files.\n\nThe dataset is correctly added to the kernel, but all the .pth files have a question mark symbol.\nWhen I click the question mark, I can read: .pth is not a supported format.\n\nThe same happens with .pkl files (pickle)\n\nAnd the same happens if I try to add it as another kernel output (instead of a dataset)\n\n\n\nI'm using Kaggle cli to create the kernel.  And I've checked that other files with extensions like .csv are working",
      "votes": null
    },
    {
      "id": "577212",
      "postDate": "07/16/2019 13:35:49",
      "content": "<p>That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now. \nAll your added datasets are at <code>../input</code>, to see run a jupyter cell with following command:\n<code>!ls ../input/</code></p>",
      "rawMarkdown": "That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now. \nAll your added datasets are at `../input`, to see run a jupyter cell with following command:\n`!ls ../input/`",
      "votes": null
    },
    {
      "id": "577215",
      "postDate": "07/16/2019 13:39:44",
      "content": "<p>Are you using the web interface or the Kaggle cli, <a href=\"/zxyu1995\">@zxyu1995</a> ?</p>",
      "rawMarkdown": "Are you using the web interface or the Kaggle cli, @zxyu1995 ?",
      "votes": null
    },
    {
      "id": "577239",
      "postDate": "07/16/2019 14:27:47",
      "content": "<p>It does the same with .hdf5 files, but you can still reference them in your kernel code.</p>",
      "rawMarkdown": "It does the same with .hdf5 files, but you can still reference them in your kernel code.",
      "votes": null
    },
    {
      "id": "577424",
      "postDate": "07/16/2019 17:08:10",
      "content": "<p>Thanks a lot <a href=\"/rishabhiitbhu\">@rishabhiitbhu</a>, <a href=\"/zxyu1995\">@zxyu1995</a> and <a href=\"/tricky999\">@tricky999</a> </p>\n\n<p>There was a typo in my paths as you indicated.</p>\n\n<p>Now, I know that the question mark symbol had nothing to do.</p>\n\n<p>Thanks!</p>",
      "rawMarkdown": "Thanks a lot @rishabhiitbhu, @zxyu1995 and @tricky999 \n\nThere was a typo in my paths as you indicated.\n\nNow, I know that the question mark symbol had nothing to do.\n\nThanks!",
      "votes": null
    },
    {
      "id": "578594",
      "postDate": "07/17/2019 22:30:04",
      "content": "<p>In my case, the .pth files have an exclamation mark symbol. \nError message is \"An error occurred parsing this file: Invalid file type: pth.\" </p>\n\n<p>The same happens with .pt and .bin files.\nIs anyone else facing this same problem as well?</p>",
      "rawMarkdown": "In my case, the .pth files have an exclamation mark symbol. \nError message is \"An error occurred parsing this file: Invalid file type: pth.\" \n\nThe same happens with .pt and .bin files.\nIs anyone else facing this same problem as well?",
      "votes": null
    },
    {
      "id": "578872",
      "postDate": "07/18/2019 08:31:10",
      "content": "<p>Hey, It's not a problem.  Copy pasting my previous comment.</p>\n\n<blockquote>\n  <p>That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now.\n  All your added datasets are at ../input, to see run a jupyter cell with following command:\n  !ls ../input/</p>\n</blockquote>",
      "rawMarkdown": "Hey, It's not a problem.  Copy pasting my previous comment.\n&gt;That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now.\nAll your added datasets are at ../input, to see run a jupyter cell with following command:\n!ls ../input/",
      "votes": null
    },
    {
      "id": "580377",
      "postDate": "07/20/2019 04:38:52",
      "content": "<p>Thank you for your advice. The problem is solved.</p>",
      "rawMarkdown": "Thank you for your advice. The problem is solved.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 577179,
      "author_name": "zxyu1995",
      "author_url": "",
      "post_date": "07/16/2019 12:53:29",
      "content": "<p>No</p>",
      "votes": null,
      "replies": [
        {
          "id": 577215,
          "author_name": "virilo",
          "author_url": "",
          "post_date": "07/16/2019 13:39:44",
          "content": "<p>Are you using the web interface or the Kaggle cli, <a href=\"/zxyu1995\">@zxyu1995</a> ?</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 577190,
      "author_name": "rishabhiitbhu",
      "author_url": "",
      "post_date": "07/16/2019 13:15:18",
      "content": "<p>Hey, can you explain specifically what problem are you facing?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 577205,
      "author_name": "virilo",
      "author_url": "",
      "post_date": "07/16/2019 13:29:18",
      "content": "<p>I added this kaggle.com/hmendonca/efficientnet-pytorch datasource to my kernel.</p>\n\n<p>It has some .pth (pytorch checkpoint) files.</p>\n\n<p>The dataset is correctly added to the kernel, but all the .pth files have a question mark symbol.\nWhen I click the question mark, I can read: .pth is not a supported format.</p>\n\n<p>The same happens with .pkl files (pickle)</p>\n\n<p>And the same happens if I try to add it as another kernel output (instead of a dataset)</p>\n\n<p>I'm using Kaggle cli to create the kernel.  And I've checked that other files with extensions like .csv are working</p>",
      "votes": null,
      "replies": [
        {
          "id": 577212,
          "author_name": "rishabhiitbhu",
          "author_url": "",
          "post_date": "07/16/2019 13:35:49",
          "content": "<p>That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now. \nAll your added datasets are at <code>../input</code>, to see run a jupyter cell with following command:\n<code>!ls ../input/</code></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 577239,
          "author_name": "tricky999",
          "author_url": "",
          "post_date": "07/16/2019 14:27:47",
          "content": "<p>It does the same with .hdf5 files, but you can still reference them in your kernel code.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 577424,
      "author_name": "virilo",
      "author_url": "",
      "post_date": "07/16/2019 17:08:10",
      "content": "<p>Thanks a lot <a href=\"/rishabhiitbhu\">@rishabhiitbhu</a>, <a href=\"/zxyu1995\">@zxyu1995</a> and <a href=\"/tricky999\">@tricky999</a> </p>\n\n<p>There was a typo in my paths as you indicated.</p>\n\n<p>Now, I know that the question mark symbol had nothing to do.</p>\n\n<p>Thanks!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 578594,
      "author_name": "yuyu12",
      "author_url": "",
      "post_date": "07/17/2019 22:30:04",
      "content": "<p>In my case, the .pth files have an exclamation mark symbol. \nError message is \"An error occurred parsing this file: Invalid file type: pth.\" </p>\n\n<p>The same happens with .pt and .bin files.\nIs anyone else facing this same problem as well?</p>",
      "votes": null,
      "replies": [
        {
          "id": 578872,
          "author_name": "rishabhiitbhu",
          "author_url": "",
          "post_date": "07/18/2019 08:31:10",
          "content": "<p>Hey, It's not a problem.  Copy pasting my previous comment.</p>\n\n<blockquote>\n  <p>That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now.\n  All your added datasets are at ../input, to see run a jupyter cell with following command:\n  !ls ../input/</p>\n</blockquote>",
          "votes": null,
          "replies": []
        },
        {
          "id": 580377,
          "author_name": "yuyu12",
          "author_url": "",
          "post_date": "07/20/2019 04:38:52",
          "content": "<p>Thank you for your advice. The problem is solved.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "577164": "are you experiencing problems adding .pth files to your kernels?\n\nhttps://github.com/Kaggle/kaggle-api/issues/202\n\nwithout using pretrained checkpoints, it's quite difficult to participate in image related competitions",
    "577179": "No",
    "577190": "Hey, can you explain specifically what problem are you facing?",
    "577205": "I added this kaggle.com/hmendonca/efficientnet-pytorch datasource to my kernel.\n\nIt has some .pth (pytorch checkpoint) files.\n\nThe dataset is correctly added to the kernel, but all the .pth files have a question mark symbol.\nWhen I click the question mark, I can read: .pth is not a supported format.\n\nThe same happens with .pkl files (pickle)\n\nAnd the same happens if I try to add it as another kernel output (instead of a dataset)\n\n\n\nI'm using Kaggle cli to create the kernel.  And I've checked that other files with extensions like .csv are working",
    "577212": "That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now. \nAll your added datasets are at `../input`, to see run a jupyter cell with following command:\n`!ls ../input/`",
    "577215": "Are you using the web interface or the Kaggle cli, @zxyu1995 ?",
    "577239": "It does the same with .hdf5 files, but you can still reference them in your kernel code.",
    "577424": "Thanks a lot @rishabhiitbhu, @zxyu1995 and @tricky999 \n\nThere was a typo in my paths as you indicated.\n\nNow, I know that the question mark symbol had nothing to do.\n\nThanks!",
    "578594": "In my case, the .pth files have an exclamation mark symbol. \nError message is \"An error occurred parsing this file: Invalid file type: pth.\" \n\nThe same happens with .pt and .bin files.\nIs anyone else facing this same problem as well?",
    "578872": "Hey, It's not a problem.  Copy pasting my previous comment.\n&gt;That message which says \".pth is not supported format\" means that it can not be interpreted and shown to you as like say a text file. That doesn't mean that you can't use that .pth file in your kernel. You can still use that file. I hope you get it now.\nAll your added datasets are at ../input, to see run a jupyter cell with following command:\n!ls ../input/",
    "580377": "Thank you for your advice. The problem is solved."
  },
  "source": "meta"
}