{
  "id": 159992,
  "title": "How to test inference code in kernel?",
  "url": "/competitions/birdsong-recognition/discussion/159992",
  "author_name": "",
  "post_date": "2020-06-19T11:45:42.412170700Z",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hello everyone! I want to thank the host for hosting such an interesting competition. I have never worked with audio data before.</p>\n\n<p>I have been able to successfully build a training and inference pipeline, but I do not know how to test my inference and submission code in the kernel as we do not have access to the full testing data. Does anyone have any clever ideas or hacks to test inference code in the kernel?</p>",
  "messages": [
    {
      "id": "893101",
      "postDate": "06/19/2020 11:45:42",
      "content": "<p>Hello everyone! I want to thank the host for hosting such an interesting competition. I have never worked with audio data before.</p>\n\n<p>I have been able to successfully build a training and inference pipeline, but I do not know how to test my inference and submission code in the kernel as we do not have access to the full testing data. Does anyone have any clever ideas or hacks to test inference code in the kernel?</p>",
      "rawMarkdown": "Hello everyone! I want to thank the host for hosting such an interesting competition. I have never worked with audio data before.\n\nI have been able to successfully build a training and inference pipeline, but I do not know how to test my inference and submission code in the kernel as we do not have access to the full testing data. Does anyone have any clever ideas or hacks to test inference code in the kernel?",
      "votes": null
    },
    {
      "id": "893104",
      "postDate": "06/19/2020 11:47:34",
      "content": "<p>I was thinking of building a dummy test dataframe using the training audio files and metadata. It will use the format of the testing dataframe, but borrow data points from the training dataframe.</p>",
      "rawMarkdown": "I was thinking of building a dummy test dataframe using the training audio files and metadata. It will use the format of the testing dataframe, but borrow data points from the training dataframe.",
      "votes": null
    },
    {
      "id": "893116",
      "postDate": "06/19/2020 12:02:18",
      "content": "<p>Yes, solution:\n<a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/159993\">https://www.kaggle.com/c/birdsong-recognition/discussion/159993</a></p>",
      "rawMarkdown": "Yes, solution:\nhttps://www.kaggle.com/c/birdsong-recognition/discussion/159993",
      "votes": null
    },
    {
      "id": "893242",
      "postDate": "06/19/2020 13:41:17",
      "content": "<p>Thank you so much, <a href=\"/shonenkov\">@shonenkov</a>! This fixed my problem. I highly appreciate your sharing :D</p>",
      "rawMarkdown": "Thank you so much, @shonenkov! This fixed my problem. I highly appreciate your sharing :D",
      "votes": null
    },
    {
      "id": "893244",
      "postDate": "06/19/2020 13:42:57",
      "content": "<p>It would be really helpful if the files and folders names were the same for the interactive and submission parts.  </p>\n\n<p>They should change \"example_test_audio\" to \"test_audio\".  That way our interactive code will work.  When we submit, swap out the dummy files and folders for the real ones.   </p>",
      "rawMarkdown": "It would be really helpful if the files and folders names were the same for the interactive and submission parts.  \n\nThey should change \"example_test_audio\" to \"test_audio\".  That way our interactive code will work.  When we submit, swap out the dummy files and folders for the real ones.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 893104,
      "author_name": "tarunpaparaju",
      "author_url": "",
      "post_date": "06/19/2020 11:47:34",
      "content": "<p>I was thinking of building a dummy test dataframe using the training audio files and metadata. It will use the format of the testing dataframe, but borrow data points from the training dataframe.</p>",
      "votes": null,
      "replies": [
        {
          "id": 893244,
          "author_name": "ericfreeman",
          "author_url": "",
          "post_date": "06/19/2020 13:42:57",
          "content": "<p>It would be really helpful if the files and folders names were the same for the interactive and submission parts.  </p>\n\n<p>They should change \"example_test_audio\" to \"test_audio\".  That way our interactive code will work.  When we submit, swap out the dummy files and folders for the real ones.   </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 893116,
      "author_name": "shonenkov",
      "author_url": "",
      "post_date": "06/19/2020 12:02:18",
      "content": "<p>Yes, solution:\n<a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/159993\">https://www.kaggle.com/c/birdsong-recognition/discussion/159993</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 893242,
          "author_name": "tarunpaparaju",
          "author_url": "",
          "post_date": "06/19/2020 13:41:17",
          "content": "<p>Thank you so much, <a href=\"/shonenkov\">@shonenkov</a>! This fixed my problem. I highly appreciate your sharing :D</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "893101": "Hello everyone! I want to thank the host for hosting such an interesting competition. I have never worked with audio data before.\n\nI have been able to successfully build a training and inference pipeline, but I do not know how to test my inference and submission code in the kernel as we do not have access to the full testing data. Does anyone have any clever ideas or hacks to test inference code in the kernel?",
    "893104": "I was thinking of building a dummy test dataframe using the training audio files and metadata. It will use the format of the testing dataframe, but borrow data points from the training dataframe.",
    "893116": "Yes, solution:\nhttps://www.kaggle.com/c/birdsong-recognition/discussion/159993",
    "893242": "Thank you so much, @shonenkov! This fixed my problem. I highly appreciate your sharing :D",
    "893244": "It would be really helpful if the files and folders names were the same for the interactive and submission parts.  \n\nThey should change \"example_test_audio\" to \"test_audio\".  That way our interactive code will work.  When we submit, swap out the dummy files and folders for the real ones."
  },
  "source": "meta"
}