{
  "id": 124401,
  "title": "Wired thing happening",
  "url": "/competitions/bengaliai-cv19/discussion/124401",
  "author_name": "Bibhas Mondal",
  "post_date": "2020-01-03T22:39:05.513000",
  "votes": -4,
  "comment_count": 8,
  "views": 0,
  "content": "<p>While submitting a submission.csv from high scoring karnel, i am getting only 0.0614 LB.</p>",
  "messages": [
    {
      "id": 709778,
      "postDate": "2020-01-03T23:17:27.990Z",
      "content": "<p>That is the score of the sample_submission.csv file (if you submit it without changing the values)</p>",
      "rawMarkdown": "That is the score of the sample_submission.csv file (if you submit it without changing the values)",
      "votes": 2,
      "replies": [
        {
          "id": 709889,
          "postDate": "2020-01-04T03:38:35.620Z",
          "content": "<p>I haven't submitted sample_submission.csv. First i have submitted my own solution but got 0.0614. After that i submitted high scoring submission.csv by making a dataset but got 0.0615, though the lb of that file was 0.9641</p>",
          "rawMarkdown": "I haven't submitted sample_submission.csv. First i have submitted my own solution but got 0.0614. After that i submitted high scoring submission.csv by making a dataset but got 0.0615, though the lb of that file was 0.9641"
        }
      ]
    },
    {
      "id": 709762,
      "postDate": "2020-01-03T22:39:05.513Z",
      "content": "<p>While submitting a submission.csv from high scoring karnel, i am getting only 0.0614 LB.</p>",
      "rawMarkdown": "While submitting a submission.csv from high scoring karnel, i am getting only 0.0614 LB.",
      "votes": -4
    },
    {
      "id": 710282,
      "postDate": "2020-01-04T14:44:10.343Z",
      "content": "<p>submission = pd.read_csv('/kaggle/input/bengaliai-cv19/sample_submission.csv')\nsubmission.set_index('row_id', inplace=True)</p>\n\n<p>predicted = pd.read_csv('/kaggle/input/dataset/submission(0.964).csv')\npredicted .set_index('row_id', inplace=True)\nfor name, row in predicted .iterrows():\n    submission.loc[name] = row</p>\n\n<p>submission.reset_index(inplace=True)\nsubmission.to_csv('submission.csv', index=False)</p>\n\n<p>Getting LB 0.0614 while submitting the submission file of LB 0.964.\nWhy this happening? Could anyone tell?</p>\n\n<p>Please help!</p>",
      "rawMarkdown": "submission = pd.read_csv('/kaggle/input/bengaliai-cv19/sample_submission.csv')\nsubmission.set_index('row_id', inplace=True)\n\npredicted = pd.read_csv('/kaggle/input/dataset/submission(0.964).csv')\npredicted .set_index('row_id', inplace=True)\nfor name, row in predicted .iterrows():\n    submission.loc[name] = row\n\nsubmission.reset_index(inplace=True)\nsubmission.to_csv('submission.csv', index=False)\n\nGetting LB 0.0614 while submitting the submission file of LB 0.964.\nWhy this happening? Could anyone tell?\n\nPlease help!",
      "replies": [
        {
          "id": 710316,
          "postDate": "2020-01-04T15:28:30.893Z",
          "content": "<p>In this competition, the sample submission file only has 12 samples (36 predictions). You can only upload predictions for those samples to a dataset. In the private re-run the sample_submission.csv contains ~200.000 records. Your script is working, but it only updates 36 values. </p>",
          "rawMarkdown": "In this competition, the sample submission file only has 12 samples (36 predictions). You can only upload predictions for those samples to a dataset. In the private re-run the sample_submission.csv contains ~200.000 records. Your script is working, but it only updates 36 values. ",
          "votes": 2
        },
        {
          "id": 710341,
          "postDate": "2020-01-04T16:03:31.490Z",
          "content": "<p>so what would be the solution?</p>",
          "rawMarkdown": "so what would be the solution?"
        },
        {
          "id": 710358,
          "postDate": "2020-01-04T16:30:11.037Z",
          "content": "<p>Do not upload csv. You can upload locally trained model's weights and predict during submission. Check the inference notebooks for ideas.</p>",
          "rawMarkdown": "Do not upload csv. You can upload locally trained model's weights and predict during submission. Check the inference notebooks for ideas.",
          "votes": 1
        },
        {
          "id": 710425,
          "postDate": "2020-01-04T17:56:36.297Z",
          "content": "<p>Thank u for your suggestion.</p>",
          "rawMarkdown": "Thank u for your suggestion.",
          "votes": 1
        }
      ]
    },
    {
      "id": 710339,
      "postDate": "2020-01-04T16:02:13.933Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 709778,
      "author_name": "Peter",
      "author_url": "",
      "post_date": "2020-01-03T23:17:27.990000",
      "content": "<p>That is the score of the sample_submission.csv file (if you submit it without changing the values)</p>",
      "votes": 2,
      "replies": [
        {
          "id": 709889,
          "author_name": "Bibhas Mondal",
          "author_url": "",
          "post_date": "2020-01-04T03:38:35.620000",
          "content": "<p>I haven't submitted sample_submission.csv. First i have submitted my own solution but got 0.0614. After that i submitted high scoring submission.csv by making a dataset but got 0.0615, though the lb of that file was 0.9641</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 710282,
      "author_name": "Bibhas Mondal",
      "author_url": "",
      "post_date": "2020-01-04T14:44:10.343000",
      "content": "<p>submission = pd.read_csv('/kaggle/input/bengaliai-cv19/sample_submission.csv')\nsubmission.set_index('row_id', inplace=True)</p>\n\n<p>predicted = pd.read_csv('/kaggle/input/dataset/submission(0.964).csv')\npredicted .set_index('row_id', inplace=True)\nfor name, row in predicted .iterrows():\n    submission.loc[name] = row</p>\n\n<p>submission.reset_index(inplace=True)\nsubmission.to_csv('submission.csv', index=False)</p>\n\n<p>Getting LB 0.0614 while submitting the submission file of LB 0.964.\nWhy this happening? Could anyone tell?</p>\n\n<p>Please help!</p>",
      "votes": 0,
      "replies": [
        {
          "id": 710316,
          "author_name": "Peter",
          "author_url": "",
          "post_date": "2020-01-04T15:28:30.893000",
          "content": "<p>In this competition, the sample submission file only has 12 samples (36 predictions). You can only upload predictions for those samples to a dataset. In the private re-run the sample_submission.csv contains ~200.000 records. Your script is working, but it only updates 36 values. </p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 710341,
          "author_name": "Bibhas Mondal",
          "author_url": "",
          "post_date": "2020-01-04T16:03:31.490000",
          "content": "<p>so what would be the solution?</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 710358,
          "author_name": "Peter",
          "author_url": "",
          "post_date": "2020-01-04T16:30:11.037000",
          "content": "<p>Do not upload csv. You can upload locally trained model's weights and predict during submission. Check the inference notebooks for ideas.</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 710425,
          "author_name": "Bibhas Mondal",
          "author_url": "",
          "post_date": "2020-01-04T17:56:36.297000",
          "content": "<p>Thank u for your suggestion.</p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 710339,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-01-04T16:02:13.933000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "709778": "That is the score of the sample_submission.csv file (if you submit it without changing the values)",
    "709762": "While submitting a submission.csv from high scoring karnel, i am getting only 0.0614 LB.",
    "710282": "submission = pd.read_csv('/kaggle/input/bengaliai-cv19/sample_submission.csv')\nsubmission.set_index('row_id', inplace=True)\n\npredicted = pd.read_csv('/kaggle/input/dataset/submission(0.964).csv')\npredicted .set_index('row_id', inplace=True)\nfor name, row in predicted .iterrows():\n    submission.loc[name] = row\n\nsubmission.reset_index(inplace=True)\nsubmission.to_csv('submission.csv', index=False)\n\nGetting LB 0.0614 while submitting the submission file of LB 0.964.\nWhy this happening? Could anyone tell?\n\nPlease help!",
    "710339": ""
  }
}