{ "id": 505614, "title": "NEED UR HELP - GUYS (SUBMISSION ERRORS)", "url": "/competitions/home-credit-credit-risk-model-stability/discussion/505614", "author_name": "", "post_date": "2024-05-18T08:43:34.583398500Z", "votes": null, "comment_count": 8, "views": 0, "content": "
    \n
  1. What is \"submission scoring error\"?
    2. \n
  2. Why does the test set takes more time than the train set?
    3. \n
", "messages": [ { "id": "2821779", "postDate": "05/18/2024 08:43:34", "content": "
    \n
  1. What is \"submission scoring error\"?
    2. \n
  2. Why does the test set takes more time than the train set?
    3. \n
", "rawMarkdown": "1. What is \"submission scoring error\"?\n2. Why does the test set takes more time than the train set?", "votes": null }, { "id": "2821986", "postDate": "05/18/2024 11:06:30", "content": "
    \n
  1. may be out of memory
    2. \n
  2. online test set is bigger than train set
    3. \n
", "rawMarkdown": "1. may be out of memory\n2. online test set is bigger than train set", "votes": null }, { "id": "2821995", "postDate": "05/18/2024 11:12:40", "content": "

thank u.

\n
    \n
  1. But when i run it with both GPU on & off it wont take like half of the available ram. i trained the model on colab and just added it here as an input.
    2. \n
  2. is it as 3X bigger than the train set? cause its taking that much time.
    3. \n
", "rawMarkdown": "thank u. \n1. But when i run it with both GPU on & off it wont take like half of the available ram. i trained the model on colab and just added it here as an input. \n2. is it as 3X bigger than the train set? cause its taking that much time.", "votes": null }, { "id": "2822000", "postDate": "05/18/2024 11:16:20", "content": "
    \n
  1. What is \"submission scoring error\"? Means that your notebook ran succesfuly and produced a submission.csv, but error occurs while scoring. Make sure you don't have NaN or null and score is float between 0-1
    2. \n
  2. the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv
    3. \n
", "rawMarkdown": "1. What is \"submission scoring error\"? Means that your notebook ran succesfuly and produced a submission.csv, but error occurs while scoring. Make sure you don't have NaN or null and score is float between 0-1\n2. the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv", "votes": null }, { "id": "2822620", "postDate": "05/18/2024 17:21:38", "content": "

ohhh, do u mean the pred values shall be float bn 0 & 1, not just 0 & 1. i converted all predictions into 0 and 1.

", "rawMarkdown": "ohhh, do u mean the pred values shall be float bn 0 & 1, not just 0 & 1. i converted all predictions into 0 and 1.", "votes": null }, { "id": "2823018", "postDate": "05/19/2024 01:02:04", "content": "

yes all value should be float, you can look at starter notebook https://www.kaggle.com/code/jetakow/home-credit-2024-starter-notebook/output

", "rawMarkdown": "yes all value should be float, you can look at starter notebook https://www.kaggle.com/code/jetakow/home-credit-2024-starter-notebook/output", "votes": null }, { "id": "2823080", "postDate": "05/19/2024 03:09:44", "content": "

thank you, sir

", "rawMarkdown": "thank you, sir", "votes": null }, { "id": "2837100", "postDate": "05/26/2024 09:18:17", "content": "

Have you fixed this issue? I got the same problem and this is the first time for global Kaggle competition.
\nI trained model and feature engineering in Colab , saving file submission as csv in colab, then upload it in Kaggle.
\nI make a few lines of code for submission and now I have submission.csv with index 0-9 in kaggle output.

\n

Submission scoring error x3 times

", "rawMarkdown": "Have you fixed this issue? I got the same problem and this is the first time for global Kaggle competition.\nI trained model and feature engineering in Colab , saving file submission as csv in colab, then upload it in Kaggle.\nI make a few lines of code for submission and now I have submission.csv with index 0-9 in kaggle output.\n\nSubmission scoring error x3 times", "votes": null }, { "id": "2837119", "postDate": "05/26/2024 09:28:40", "content": "

make sure ur submission.csv has \"case_id\" and \"score\" columns. That was my problem.

", "rawMarkdown": "make sure ur submission.csv has \"case_id\" and \"score\" columns. That was my problem.", "votes": null } ], "comments": [ { "id": 2821986, "author_name": "abc1705", "author_url": "", "post_date": "05/18/2024 11:06:30", "content": "
    \n
  1. may be out of memory
    2. \n
  2. online test set is bigger than train set
    3. \n
", "votes": null, "replies": [ { "id": 2821995, "author_name": "turebogale", "author_url": "", "post_date": "05/18/2024 11:12:40", "content": "

thank u.

\n
    \n
  1. But when i run it with both GPU on & off it wont take like half of the available ram. i trained the model on colab and just added it here as an input.
    2. \n
  2. is it as 3X bigger than the train set? cause its taking that much time.
    3. \n
", "votes": null, "replies": [ { "id": 2822000, "author_name": "waechter", "author_url": "", "post_date": "05/18/2024 11:16:20", "content": "
    \n
  1. What is \"submission scoring error\"? Means that your notebook ran succesfuly and produced a submission.csv, but error occurs while scoring. Make sure you don't have NaN or null and score is float between 0-1
    2. \n
  2. the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv
    3. \n
", "votes": null, "replies": [ { "id": 2822620, "author_name": "turebogale", "author_url": "", "post_date": "05/18/2024 17:21:38", "content": "

ohhh, do u mean the pred values shall be float bn 0 & 1, not just 0 & 1. i converted all predictions into 0 and 1.

", "votes": null, "replies": [ { "id": 2823018, "author_name": "waechter", "author_url": "", "post_date": "05/19/2024 01:02:04", "content": "

yes all value should be float, you can look at starter notebook https://www.kaggle.com/code/jetakow/home-credit-2024-starter-notebook/output

", "votes": null, "replies": [ { "id": 2823080, "author_name": "turebogale", "author_url": "", "post_date": "05/19/2024 03:09:44", "content": "

thank you, sir

", "votes": null, "replies": [] } ] } ] } ] } ] } ] }, { "id": 2837100, "author_name": "natdanaisangpho", "author_url": "", "post_date": "05/26/2024 09:18:17", "content": "

Have you fixed this issue? I got the same problem and this is the first time for global Kaggle competition.
\nI trained model and feature engineering in Colab , saving file submission as csv in colab, then upload it in Kaggle.
\nI make a few lines of code for submission and now I have submission.csv with index 0-9 in kaggle output.

\n

Submission scoring error x3 times

", "votes": null, "replies": [ { "id": 2837119, "author_name": "turebogale", "author_url": "", "post_date": "05/26/2024 09:28:40", "content": "

make sure ur submission.csv has \"case_id\" and \"score\" columns. That was my problem.

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2821779": "1. What is \"submission scoring error\"?\n2. Why does the test set takes more time than the train set?", "2821986": "1. may be out of memory\n2. online test set is bigger than train set", "2821995": "thank u. \n1. But when i run it with both GPU on & off it wont take like half of the available ram. i trained the model on colab and just added it here as an input. \n2. is it as 3X bigger than the train set? cause its taking that much time.", "2822000": "1. What is \"submission scoring error\"? Means that your notebook ran succesfuly and produced a submission.csv, but error occurs while scoring. Make sure you don't have NaN or null and score is float between 0-1\n2. the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv", "2822620": "ohhh, do u mean the pred values shall be float bn 0 & 1, not just 0 & 1. i converted all predictions into 0 and 1.", "2823018": "yes all value should be float, you can look at starter notebook https://www.kaggle.com/code/jetakow/home-credit-2024-starter-notebook/output", "2823080": "thank you, sir", "2837100": "Have you fixed this issue? I got the same problem and this is the first time for global Kaggle competition.\nI trained model and feature engineering in Colab , saving file submission as csv in colab, then upload it in Kaggle.\nI make a few lines of code for submission and now I have submission.csv with index 0-9 in kaggle output.\n\nSubmission scoring error x3 times", "2837119": "make sure ur submission.csv has \"case_id\" and \"score\" columns. That was my problem." }, "source": "meta" }