![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F4283293%2F4edd7e887c6670638543e07ae647f9a0%2FCapture.PNG?generation=1578806478378957&alt=media\")
.I managed to train a model offline and have got it loaded into my kernel (using \"pickle\" module). My kernel do successfully commit and run on test_imgs parquets. But after submitting, i am getting \"Submission Scoring Error\".
\n\nAlso, instead of passing testimages parquet files, i loaded the trainimages parquet file, just to check whether my notebook will work for similar number of private testimgs parquet dataset ( ~ 200K samples). Kernel worked sucessfully. This ensures that there are no computing/memory issues. ( This was done just for verification, i am using testimgs parquets for submission).
\n\nCan anyone help me what I am missing in my kernel. Following is the link to my notebook.\nhttps://www.kaggle.com/ankitsajwan/kernelc4d0c4d19c
", "messages": [ { "id": "716319", "postDate": "01/11/2020 14:47:58", "content": "I managed to train a model offline and have got it loaded into my kernel (using \"pickle\" module). My kernel do successfully commit and run on test_imgs parquets. But after submitting, i am getting \"Submission Scoring Error\".
\n\nAlso, instead of passing testimages parquet files, i loaded the trainimages parquet file, just to check whether my notebook will work for similar number of private testimgs parquet dataset ( ~ 200K samples). Kernel worked sucessfully. This ensures that there are no computing/memory issues. ( This was done just for verification, i am using testimgs parquets for submission).
\n\nCan anyone help me what I am missing in my kernel. Following is the link to my notebook.\nhttps://www.kaggle.com/ankitsajwan/kernelc4d0c4d19c
", "rawMarkdown": "I managed to train a model offline and have got it loaded into my kernel (using \"pickle\" module). My kernel do successfully commit and run on test_imgs parquets. But after submitting, i am getting \"**Submission Scoring Error**\".\n\nAlso, instead of passing testimages parquet files, i loaded the trainimages parquet file, just to check whether my notebook will work for similar number of private testimgs parquet dataset ( ~ **200K samples**). Kernel worked sucessfully. This ensures that there are no computing/memory issues. ( This was done just for verification, i am using testimgs parquets for submission).\n\nCan anyone help me what I am missing in my kernel. Following is the link to my notebook.\nhttps://www.kaggle.com/ankitsajwan/kernelc4d0c4d19c", "votes": null }, { "id": "716547", "postDate": "01/11/2020 21:06:59", "content": "Your notebook link seems to raise a 404 HTTP Error code when following the link, maybe it is private ?
", "rawMarkdown": "Your notebook link seems to raise a 404 HTTP Error code when following the link, maybe it is private ?", "votes": null }, { "id": "716693", "postDate": "01/12/2020 05:24:27", "content": "Hi @dimartinot. I have made the notebook private as I have resolved the issue.
\n\nI made changes on how i was generating the row_id column for output submission.csv.
\n\n.
Change I did was to use the imageid column from test_imgs parquet to generate row_id column instead of doing the above.
", "rawMarkdown": "Hi @dimartinot. I have made the notebook private as I have resolved the issue.\n\nI made changes on how i was generating the row_id column for output submission.csv.\n\n.\n\n**Change I did was to use the imageid column from test_imgs parquet to generate row_id column instead of doing the above.**", "votes": null } ], "comments": [ { "id": 716547, "author_name": "dimartinot", "author_url": "", "post_date": "01/11/2020 21:06:59", "content": "Your notebook link seems to raise a 404 HTTP Error code when following the link, maybe it is private ?
", "votes": null, "replies": [] }, { "id": 716693, "author_name": "ankitsajwan", "author_url": "", "post_date": "01/12/2020 05:24:27", "content": "Hi @dimartinot. I have made the notebook private as I have resolved the issue.
\n\nI made changes on how i was generating the row_id column for output submission.csv.
\n\n.
Change I did was to use the imageid column from test_imgs parquet to generate row_id column instead of doing the above.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "716319": "I managed to train a model offline and have got it loaded into my kernel (using \"pickle\" module). My kernel do successfully commit and run on test_imgs parquets. But after submitting, i am getting \"**Submission Scoring Error**\".\n\nAlso, instead of passing testimages parquet files, i loaded the trainimages parquet file, just to check whether my notebook will work for similar number of private testimgs parquet dataset ( ~ **200K samples**). Kernel worked sucessfully. This ensures that there are no computing/memory issues. ( This was done just for verification, i am using testimgs parquets for submission).\n\nCan anyone help me what I am missing in my kernel. Following is the link to my notebook.\nhttps://www.kaggle.com/ankitsajwan/kernelc4d0c4d19c", "716547": "Your notebook link seems to raise a 404 HTTP Error code when following the link, maybe it is private ?", "716693": "Hi @dimartinot. I have made the notebook private as I have resolved the issue.\n\nI made changes on how i was generating the row_id column for output submission.csv.\n\n.\n\n**Change I did was to use the imageid column from test_imgs parquet to generate row_id column instead of doing the above.**" }, "source": "meta" }