I am getting \"Submission Scoring Error\" on every submission, no matter what.
\nYet, the submission.csv loads fine when I open it and seems to show reasonable results when overlayed over the images. I am using mask2rle and rle2mask functions published in a number of notebooks by the other participants.
\nWhat are the validation rules and can we have them spelled out?
\nThis is very frustrating.
I am getting \"Submission Scoring Error\" on every submission, no matter what.
\nYet, the submission.csv loads fine when I open it and seems to show reasonable results when overlayed over the images. I am using mask2rle and rle2mask functions published in a number of notebooks by the other participants.
\nWhat are the validation rules and can we have them spelled out?
\nThis is very frustrating.
I am having the same issue, how you have solved?
", "rawMarkdown": "I am having the same issue, how you have solved?", "votes": null }, { "id": "1100226", "postDate": "12/02/2020 23:29:10", "content": "As a matter of fact yes - I was able to make a successful submission today.
\nI think there are a couple points missing from the official description:
Hope this helps!
", "rawMarkdown": "As a matter of fact yes - I was able to make a successful submission today.\nI think there are a couple points missing from the official description:\n1. Do not use hardcoded image ids or filenames. The submission seems to be scored on more files than is available when you run the notebook in edit mode. Use os.walk() or a similar function to loop over the available test image files.\n2. The submission takes much longer than one would expect. I can run the inference on the whole test file set in 15 minutes on a CPU-based instance, but the submission took more than an hour. So it is very easy to get a timeout error, you may want to optimize the inference as much as you can - which is challenging because the RAM available to the instance is rather puny at 16 Gb.\n\nHope this helps!", "votes": null }, { "id": "1100607", "postDate": "12/03/2020 07:28:11", "content": "Can you share the code of the submission?
", "rawMarkdown": "Can you share the code of the submission?", "votes": null }, { "id": "1100857", "postDate": "12/03/2020 12:09:57", "content": "Here:
\nhttps://www.kaggle.com/igor14497/making-a-successful-submission
I am having the same issue, how you have solved?
", "rawMarkdown": "I am having the same issue, how you have solved?", "votes": null }, { "id": "1111786", "postDate": "12/14/2020 03:42:18", "content": "See the discussion here. I commented on the solution 11 days ago in this topic. I also posted a notebook for a successful submission. Not a record-beating, but a successful submission.
", "rawMarkdown": "See the discussion here. I commented on the solution 11 days ago in this topic. I also posted a notebook for a successful submission. Not a record-beating, but a successful submission.", "votes": null }, { "id": "1111804", "postDate": "12/14/2020 03:58:53", "content": "But your noteboook has no output
", "rawMarkdown": "But your noteboook has no output", "votes": null }, { "id": "1111822", "postDate": "12/14/2020 04:15:34", "content": "I am confused by your statement.
\nThe very last line in my notebook is
\nsubmission_df.to_csv('submission.csv',index=False)
\nsubmission.csv is the output required for scoring. What other output do you mean?
\nI assure you I ran this notebook and it was scored successfully, if not very high.
Thank you . So can you tell me the reason for the normal submission?
", "rawMarkdown": "Thank you . So can you tell me the reason for the normal submission?", "votes": null }, { "id": "1111896", "postDate": "12/14/2020 05:58:09", "content": "It is explained in the second comment in this very topic. Is it so hard to read?
", "rawMarkdown": "It is explained in the second comment in this very topic. Is it so hard to read?", "votes": null }, { "id": "1112061", "postDate": "12/14/2020 08:42:34", "content": "thank you ! I see
", "rawMarkdown": "thank you ! I see", "votes": null } ], "comments": [ { "id": 1100200, "author_name": "marcelosanchezortega", "author_url": "", "post_date": "12/02/2020 22:58:13", "content": "I am having the same issue, how you have solved?
", "votes": null, "replies": [] }, { "id": 1100226, "author_name": "igor14497", "author_url": "", "post_date": "12/02/2020 23:29:10", "content": "As a matter of fact yes - I was able to make a successful submission today.
\nI think there are a couple points missing from the official description:
Hope this helps!
", "votes": null, "replies": [ { "id": 1100607, "author_name": "marcelosanchezortega", "author_url": "", "post_date": "12/03/2020 07:28:11", "content": "Can you share the code of the submission?
", "votes": null, "replies": [] } ] }, { "id": 1100857, "author_name": "igor14497", "author_url": "", "post_date": "12/03/2020 12:09:57", "content": "Here:
\nhttps://www.kaggle.com/igor14497/making-a-successful-submission
I am having the same issue, how you have solved?
", "votes": null, "replies": [ { "id": 1111786, "author_name": "igor14497", "author_url": "", "post_date": "12/14/2020 03:42:18", "content": "See the discussion here. I commented on the solution 11 days ago in this topic. I also posted a notebook for a successful submission. Not a record-beating, but a successful submission.
", "votes": null, "replies": [] }, { "id": 1111804, "author_name": "lzzkathy", "author_url": "", "post_date": "12/14/2020 03:58:53", "content": "But your noteboook has no output
", "votes": null, "replies": [] }, { "id": 1111822, "author_name": "igor14497", "author_url": "", "post_date": "12/14/2020 04:15:34", "content": "I am confused by your statement.
\nThe very last line in my notebook is
\nsubmission_df.to_csv('submission.csv',index=False)
\nsubmission.csv is the output required for scoring. What other output do you mean?
\nI assure you I ran this notebook and it was scored successfully, if not very high.
Thank you . So can you tell me the reason for the normal submission?
", "votes": null, "replies": [ { "id": 1111896, "author_name": "igor14497", "author_url": "", "post_date": "12/14/2020 05:58:09", "content": "It is explained in the second comment in this very topic. Is it so hard to read?
", "votes": null, "replies": [] } ] }, { "id": 1112061, "author_name": "lzzkathy", "author_url": "", "post_date": "12/14/2020 08:42:34", "content": "thank you ! I see
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1095513": "I am getting \"Submission Scoring Error\" on every submission, no matter what.\nYet, the submission.csv loads fine when I open it and seems to show reasonable results when overlayed over the images. I am using mask2rle and rle2mask functions published in a number of notebooks by the other participants.\nWhat are the validation rules and can we have them spelled out?\nThis is very frustrating.", "1100200": "I am having the same issue, how you have solved?", "1100226": "As a matter of fact yes - I was able to make a successful submission today.\nI think there are a couple points missing from the official description:\n1. Do not use hardcoded image ids or filenames. The submission seems to be scored on more files than is available when you run the notebook in edit mode. Use os.walk() or a similar function to loop over the available test image files.\n2. The submission takes much longer than one would expect. I can run the inference on the whole test file set in 15 minutes on a CPU-based instance, but the submission took more than an hour. So it is very easy to get a timeout error, you may want to optimize the inference as much as you can - which is challenging because the RAM available to the instance is rather puny at 16 Gb.\n\nHope this helps!", "1100607": "Can you share the code of the submission?", "1100857": "Here:\nhttps://www.kaggle.com/igor14497/making-a-successful-submission", "1111768": "I am having the same issue, how you have solved?", "1111786": "See the discussion here. I commented on the solution 11 days ago in this topic. I also posted a notebook for a successful submission. Not a record-beating, but a successful submission.", "1111804": "But your noteboook has no output", "1111822": "I am confused by your statement.\nThe very last line in my notebook is\n`submission_df.to_csv('submission.csv',index=False) `\nsubmission.csv is the output required for scoring. What other output do you mean?\nI assure you I ran this notebook and it was scored successfully, if not very high.", "1111834": "Thank you . So can you tell me the reason for the normal submission?", "1111896": "It is explained in the second comment in this very topic. Is it so hard to read?", "1112061": "thank you ! I see" }, "source": "meta" }