{
  "id": 131377,
  "title": "Submission CSV not found.",
  "url": "/competitions/deepfake-detection-challenge/discussion/131377",
  "author_name": "",
  "post_date": "2020-02-19T14:44:44.924150100Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi,</p>\n\n<p>The code is committing successfully and also outputing a submission.csv file. I tried submitting my output a several times, but every time I get this error. I am using an externally trained model. After I load the mode, I read the filenames from sample_submission.csv, store those in a list, then in the same order my model makes predictions and stores the results in a list, finally I make a dataframe which takes both the lists and writes submission.csv in the output folder (looks exactly like sample_submission). I don't understand where am I going wrong. Pls help!</p>",
  "messages": [
    {
      "id": "750614",
      "postDate": "02/19/2020 14:44:44",
      "content": "<p>Hi,</p>\n\n<p>The code is committing successfully and also outputing a submission.csv file. I tried submitting my output a several times, but every time I get this error. I am using an externally trained model. After I load the mode, I read the filenames from sample_submission.csv, store those in a list, then in the same order my model makes predictions and stores the results in a list, finally I make a dataframe which takes both the lists and writes submission.csv in the output folder (looks exactly like sample_submission). I don't understand where am I going wrong. Pls help!</p>",
      "rawMarkdown": "Hi,\n\nThe code is committing successfully and also outputing a submission.csv file. I tried submitting my output a several times, but every time I get this error. I am using an externally trained model. After I load the mode, I read the filenames from sample_submission.csv, store those in a list, then in the same order my model makes predictions and stores the results in a list, finally I make a dataframe which takes both the lists and writes submission.csv in the output folder (looks exactly like sample_submission). I don't understand where am I going wrong. Pls help!",
      "votes": null
    },
    {
      "id": "750957",
      "postDate": "02/19/2020 21:44:16",
      "content": "<p>You should not be reading samplesubmission.csv at all. The Notebook should make predictions on the videos in the test_videos folder (e.g. \"/kaggle/input/deepfake-detection-challenge/test_videos/*.mp4\"). It should read all files in this folder, extract the name of the file (i.e. basename) then use your model to predict if the video is real or fake. Then write the base filename and label to the submission.csv file.</p>\n\n<p>There are a few inference kernels already. I suggest try one first and it will help you understand where you are going wrong.</p>",
      "rawMarkdown": "You should not be reading samplesubmission.csv at all. The Notebook should make predictions on the videos in the test_videos folder (e.g. \"/kaggle/input/deepfake-detection-challenge/test_videos/*.mp4\"). It should read all files in this folder, extract the name of the file (i.e. basename) then use your model to predict if the video is real or fake. Then write the base filename and label to the submission.csv file.\n\nThere are a few inference kernels already. I suggest try one first and it will help you understand where you are going wrong.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 750957,
      "author_name": "maralski",
      "author_url": "",
      "post_date": "02/19/2020 21:44:16",
      "content": "<p>You should not be reading samplesubmission.csv at all. The Notebook should make predictions on the videos in the test_videos folder (e.g. \"/kaggle/input/deepfake-detection-challenge/test_videos/*.mp4\"). It should read all files in this folder, extract the name of the file (i.e. basename) then use your model to predict if the video is real or fake. Then write the base filename and label to the submission.csv file.</p>\n\n<p>There are a few inference kernels already. I suggest try one first and it will help you understand where you are going wrong.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "750614": "Hi,\n\nThe code is committing successfully and also outputing a submission.csv file. I tried submitting my output a several times, but every time I get this error. I am using an externally trained model. After I load the mode, I read the filenames from sample_submission.csv, store those in a list, then in the same order my model makes predictions and stores the results in a list, finally I make a dataframe which takes both the lists and writes submission.csv in the output folder (looks exactly like sample_submission). I don't understand where am I going wrong. Pls help!",
    "750957": "You should not be reading samplesubmission.csv at all. The Notebook should make predictions on the videos in the test_videos folder (e.g. \"/kaggle/input/deepfake-detection-challenge/test_videos/*.mp4\"). It should read all files in this folder, extract the name of the file (i.e. basename) then use your model to predict if the video is real or fake. Then write the base filename and label to the submission.csv file.\n\nThere are a few inference kernels already. I suggest try one first and it will help you understand where you are going wrong."
  },
  "source": "meta"
}