![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F2040189%2F348944b777173f29ed5e17f61020095e%2F12.png?generation=1697789198712168&alt=media\")
Hello everyone There are a questions.
\n1) Submission Scoring Error
\nThere are 2 codes
\n1.1) I load the source data, build a model, upload the response to output. I upload this code to submission and it accepts the result and puts down a score.
\n1.2) I upload a response file to inputs (I take this file from another dataset) and upload it to outputs. That is, I do not take the initial data, I do not build a model
The second code does not accept, the first code accepts. Why is that?
\nYou cannot use the Internet in the task. But in the comments people write that you can use datasets in the solution
How do I use data from another dataset that I received in another code when loading the solution? How do I use the model I got in another program?
\n2) Notebook Out of Memory
\nThe only model that I took the competition was considered 344 seconds. All the other models that I counted, it was written \"Notebook Out of Memory\" everywhere, although I greatly reduced the data, my program counts only 43 seconds. But for some reason he writes that \"Notebook Out of Memory\" and accepts the answer for more than an hour, although the program itself is counted in Kaggle for only 43 seconds, how can this be? I used xgb.XGBClassifier in all cases.
\nWherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".
\nPlease help me, this is my first job.
\nAnd instead of doing the task for more than a week, I can't get at least 1 normal answer from morning to evening.
Hello everyone There are a questions.
\n1) Submission Scoring Error
\nThere are 2 codes
\n1.1) I load the source data, build a model, upload the response to output. I upload this code to submission and it accepts the result and puts down a score.
\n1.2) I upload a response file to inputs (I take this file from another dataset) and upload it to outputs. That is, I do not take the initial data, I do not build a model
The second code does not accept, the first code accepts. Why is that?
\nYou cannot use the Internet in the task. But in the comments people write that you can use datasets in the solution
How do I use data from another dataset that I received in another code when loading the solution? How do I use the model I got in another program?
\n2) Notebook Out of Memory
\nThe only model that I took the competition was considered 344 seconds. All the other models that I counted, it was written \"Notebook Out of Memory\" everywhere, although I greatly reduced the data, my program counts only 43 seconds. But for some reason he writes that \"Notebook Out of Memory\" and accepts the answer for more than an hour, although the program itself is counted in Kaggle for only 43 seconds, how can this be? I used xgb.XGBClassifier in all cases.
\nWherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".
\nPlease help me, this is my first job.
\nAnd instead of doing the task for more than a week, I can't get at least 1 normal answer from morning to evening.
The only answer that accepted gave a score of 0
", "rawMarkdown": "The only answer that accepted gave a score of 0", "votes": null }, { "id": "2490408", "postDate": "10/20/2023 16:36:35", "content": "Hello @rinatbayanov
\n\n\nThe second code does not accept, the first code accepts. Why is that?
\n
What is the error you are getting? Want to understand a little more on what you mean by the \"does not accept\".
\n\n\nWherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".
\n
Please go through the link below. There are some help hints on what could be wrong during submission based on the type of error you are getting and some additional guidelines to help debug errors during submissions.
\nRef: https://www.kaggle.com/code-competition-debugging
", "rawMarkdown": "Hello @rinatbayanov \n\n> The second code does not accept, the first code accepts. Why is that?\n\nWhat is the error you are getting? Want to understand a little more on what you mean by the \"does not accept\".\n\n> Wherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".\n\nPlease go through the link below. There are some help hints on what could be wrong during submission based on the type of error you are getting and some additional guidelines to help debug errors during submissions.\n\nRef: https://www.kaggle.com/code-competition-debugging", "votes": null }, { "id": "2491895", "postDate": "10/22/2023 06:04:52", "content": "I think the reason is that the output csv file you get after saving a version is not the actual result. The notebook needs to run on the hidden dataset for scoring, which means it would take a lot more time to get the lb score and you can't download the output csv files to submit or combine them directly for scoring.
", "rawMarkdown": "I think the reason is that the output csv file you get after saving a version is not the actual result. The notebook needs to run on the hidden dataset for scoring, which means it would take a lot more time to get the lb score and you can't download the output csv files to submit or combine them directly for scoring.", "votes": null }, { "id": "2497928", "postDate": "10/25/2023 02:41:01", "content": "Thank you, I didn't know there was a hidden dataset, you have a good answer
", "rawMarkdown": "Thank you, I didn't know there was a hidden dataset, you have a good answer", "votes": null }, { "id": "2497931", "postDate": "10/25/2023 02:42:52", "content": "thank you! I spent 2 weeks trying to understand that there is a hidden set of submission data, 2 weeks I sat for 8 hours a day. Now I'm rebuilding submission so that it works on any data, let's see what happens
", "rawMarkdown": "thank you! I spent 2 weeks trying to understand that there is a hidden set of submission data, 2 weeks I sat for 8 hours a day. Now I'm rebuilding submission so that it works on any data, let's see what happens", "votes": null } ], "comments": [ { "id": 2489804, "author_name": "rinatbayanov", "author_url": "", "post_date": "10/20/2023 08:21:20", "content": "The only answer that accepted gave a score of 0
", "votes": null, "replies": [] }, { "id": 2490408, "author_name": "bkowshik", "author_url": "", "post_date": "10/20/2023 16:36:35", "content": "Hello @rinatbayanov
\n\n\nThe second code does not accept, the first code accepts. Why is that?
\n
What is the error you are getting? Want to understand a little more on what you mean by the \"does not accept\".
\n\n\nWherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".
\n
Please go through the link below. There are some help hints on what could be wrong during submission based on the type of error you are getting and some additional guidelines to help debug errors during submissions.
\nRef: https://www.kaggle.com/code-competition-debugging
", "votes": null, "replies": [ { "id": 2497931, "author_name": "rinatbayanov", "author_url": "", "post_date": "10/25/2023 02:42:52", "content": "thank you! I spent 2 weeks trying to understand that there is a hidden set of submission data, 2 weeks I sat for 8 hours a day. Now I'm rebuilding submission so that it works on any data, let's see what happens
", "votes": null, "replies": [] } ] }, { "id": 2491895, "author_name": "lizhecheng", "author_url": "", "post_date": "10/22/2023 06:04:52", "content": "I think the reason is that the output csv file you get after saving a version is not the actual result. The notebook needs to run on the hidden dataset for scoring, which means it would take a lot more time to get the lb score and you can't download the output csv files to submit or combine them directly for scoring.
", "votes": null, "replies": [ { "id": 2497928, "author_name": "rinatbayanov", "author_url": "", "post_date": "10/25/2023 02:41:01", "content": "Thank you, I didn't know there was a hidden dataset, you have a good answer
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2489802": "Hello everyone There are a questions. \n\n**1) Submission Scoring Error**\n\nThere are 2 codes\n1.1) I load the source data, build a model, upload the response to output. I upload this code to submission and it accepts the result and puts down a score.\n1.2) I upload a response file to inputs (I take this file from another dataset) and upload it to outputs. That is, I do not take the initial data, I do not build a model\n\nThe second code does not accept, the first code accepts. Why is that?\nYou cannot use the Internet in the task. But in the comments people write that you can use datasets in the solution\n\n\n\nHow do I use data from another dataset that I received in another code when loading the solution? How do I use the model I got in another program?\n\n**2) Notebook Out of Memory**\n\nThe only model that I took the competition was considered 344 seconds. All the other models that I counted, it was written \"Notebook Out of Memory\" everywhere, although I greatly reduced the data, my program counts only 43 seconds. But for some reason he writes that \"Notebook Out of Memory\" and accepts the answer for more than an hour, although the program itself is counted in Kaggle for only 43 seconds, how can this be? I used xgb.XGBClassifier in all cases.\n\nWherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".\n\nPlease help me, this is my first job.\nAnd instead of doing the task for more than a week, I can't get at least 1 normal answer from morning to evening.", "2489804": "The only answer that accepted gave a score of 0", "2490408": "Hello @rinatbayanov \n\n> The second code does not accept, the first code accepts. Why is that?\n\nWhat is the error you are getting? Want to understand a little more on what you mean by the \"does not accept\".\n\n> Wherever I insert an answer from another dataset immediately, it writes \"Submission Scoring Error\". In all cases, except one where I don't use third-party models or third-party data, writes \"Notebook Out of Memory\".\n\nPlease go through the link below. There are some help hints on what could be wrong during submission based on the type of error you are getting and some additional guidelines to help debug errors during submissions.\n\nRef: https://www.kaggle.com/code-competition-debugging", "2491895": "I think the reason is that the output csv file you get after saving a version is not the actual result. The notebook needs to run on the hidden dataset for scoring, which means it would take a lot more time to get the lb score and you can't download the output csv files to submit or combine them directly for scoring.", "2497928": "Thank you, I didn't know there was a hidden dataset, you have a good answer", "2497931": "thank you! I spent 2 weeks trying to understand that there is a hidden set of submission data, 2 weeks I sat for 8 hours a day. Now I'm rebuilding submission so that it works on any data, let's see what happens" }, "source": "meta" }