![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2128949%2Fcdaf5260779f06ea3af6af6c5b4c2629%2FScreenshot%20from%202020-03-29%2019-39-24.png?generation=1585536232242457&alt=media\")
Hello - I am getting submission errors as well as some of the others on this forum.
\n\nI have some questions for the veterans out there:
\n\nThis is my first time submitting for a kaggle competition - do submissions with an error count towards my limit of 2 per day? Right now it looks like the submissions will prevent me from trying again.
My model needed to run locally since it required more RAM than kaggle notebooks allowed - do I need to get the full model to run the entire pipeline to successfully submit?
\n\n
Per suggestions in other posts I made sure that my file contains no NaN values and that outputs are written as floats. Is there anything else that I might be doing wrong? I've worked really hard on this and am very concerned that I won't be able to submit my results - someone please help!
\n\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2128949%2F39a0961813c1535c101e599fd92e059a%2FScreenshot%20from%202020-03-29%2019-47-00.png?generation=1585536441340928&alt=media\")
Hello - I am getting submission errors as well as some of the others on this forum.
\n\nI have some questions for the veterans out there:
\n\nThis is my first time submitting for a kaggle competition - do submissions with an error count towards my limit of 2 per day? Right now it looks like the submissions will prevent me from trying again.
My model needed to run locally since it required more RAM than kaggle notebooks allowed - do I need to get the full model to run the entire pipeline to successfully submit?
\n\n
Per suggestions in other posts I made sure that my file contains no NaN values and that outputs are written as floats. Is there anything else that I might be doing wrong? I've worked really hard on this and am very concerned that I won't be able to submit my results - someone please help!
\n\n
You cannot create a CSV file offline and then put it in the notebook. The video files used for the leaderboard are private, and are not the same as the ones provided to you.
", "rawMarkdown": "You cannot create a CSV file offline and then put it in the notebook. The video files used for the leaderboard are private, and are not the same as the ones provided to you.", "votes": null }, { "id": "791732", "postDate": "03/30/2020 16:07:51", "content": "My model requires > 16mb of ram to run - what are my options to submit? Can I use the kaggle API via command line to create a kernel from my PC?
", "rawMarkdown": "My model requires > 16mb of ram to run - what are my options to submit? Can I use the kaggle API via command line to create a kernel from my PC?", "votes": null }, { "id": "791944", "postDate": "03/30/2020 19:45:17", "content": "The limit is 16 GB, not MB. 😄 Assuming that is what you meant, you'll need to break up your model or reduce the amount of data you give it so that you don't go over the 16 GB limit.
\n\nThere is no way to run the kernel on your own hardware. You'll need to work within the constraints of the Kaggle notebooks.
", "rawMarkdown": "The limit is 16 GB, not MB. 😄 Assuming that is what you meant, you'll need to break up your model or reduce the amount of data you give it so that you don't go over the 16 GB limit. \n\nThere is no way to run the kernel on your own hardware. You'll need to work within the constraints of the Kaggle notebooks.", "votes": null } ], "comments": [ { "id": 791188, "author_name": "jodonghee", "author_url": "", "post_date": "03/30/2020 07:14:04", "content": "You cannot create a CSV file offline and then put it in the notebook. The video files used for the leaderboard are private, and are not the same as the ones provided to you.
", "votes": null, "replies": [ { "id": 791732, "author_name": "goalieperson", "author_url": "", "post_date": "03/30/2020 16:07:51", "content": "My model requires > 16mb of ram to run - what are my options to submit? Can I use the kaggle API via command line to create a kernel from my PC?
", "votes": null, "replies": [] }, { "id": 791944, "author_name": "humananalog", "author_url": "", "post_date": "03/30/2020 19:45:17", "content": "The limit is 16 GB, not MB. 😄 Assuming that is what you meant, you'll need to break up your model or reduce the amount of data you give it so that you don't go over the 16 GB limit.
\n\nThere is no way to run the kernel on your own hardware. You'll need to work within the constraints of the Kaggle notebooks.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "790997": "Hello - I am getting submission errors as well as some of the others on this forum.\n\nI have some questions for the veterans out there:\n\n1. This is my first time submitting for a kaggle competition - do submissions with an error count towards my limit of 2 per day? Right now it looks like the submissions will prevent me from trying again.\n\n2. My model needed to run locally since it required more RAM than kaggle notebooks allowed - do I need to get the full model to run the entire pipeline to successfully submit?\n- currently I have my submission file uploaded to my inputs, then reloaded via pandas and then finally written to my outputs folder.\n\n\n \nPer suggestions in other posts I made sure that my file contains no NaN values and that outputs are written as floats. Is there anything else that I might be doing wrong? I've worked really hard on this and am very concerned that I won't be able to submit my results - someone please help!\n\n", "791188": "1. yes\n2.You need to add the model file to 'Notebook'.\nThe error came in half an hour, and if it wasn't the intended time, the model might have stopped in the middle.", "791344": "You cannot create a CSV file offline and then put it in the notebook. The video files used for the leaderboard are private, and are not the same as the ones provided to you.", "791732": "My model requires > 16mb of ram to run - what are my options to submit? Can I use the kaggle API via command line to create a kernel from my PC?", "791944": "The limit is 16 GB, not MB. 😄 Assuming that is what you meant, you'll need to break up your model or reduce the amount of data you give it so that you don't go over the 16 GB limit. \n\nThere is no way to run the kernel on your own hardware. You'll need to work within the constraints of the Kaggle notebooks." }, "source": "meta" }