![\"Image\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F568032%2Ffe991159effba42ec133cbfa0c74b4b5%2F2024-04-02%20%209.20.13.png?generation=1712017224124757&alt=media\")
\nLog says it ran suceeded but scoring is failed
Hello, all.
\nI've made my own notebook and run everything OK with the notebook itself.
\nAnd I've submitted the notebook but I got the error of submission scoring error.
\nI checked the log and everything was fine, and I can even see the predicted / created file of sample test set.
I have no idea why this error occurs…
\nDoes any of you experiences like me?
", "messages": [ { "id": "2726530", "postDate": "04/01/2024 08:49:51", "content": "Hello, all.
\nI've made my own notebook and run everything OK with the notebook itself.
\nAnd I've submitted the notebook but I got the error of submission scoring error.
\nI checked the log and everything was fine, and I can even see the predicted / created file of sample test set.
I have no idea why this error occurs…
\nDoes any of you experiences like me?
", "rawMarkdown": "Hello, all.\n\nI've made my own notebook and run everything OK with the notebook itself.\nAnd I've submitted the notebook but I got the error of submission scoring error.\nI checked the log and everything was fine, and I can even see the predicted / created file of sample test set.\n\nI have no idea why this error occurs...\n\nDoes any of you experiences like me?", "votes": null }, { "id": "2726969", "postDate": "04/01/2024 14:06:42", "content": "Hey I had a similar bug driving me crazy a couple days ago. Basically I was generating multiple case id predictions because of bad aggregations. So make sure you are only outputting 10 rows of predictions (case id is always without duplicated).
\nGood luck!
In submissions section check the error written under the version number of failed scoring, it could be - notebook out of memory
", "rawMarkdown": "In submissions section check the error written under the version number of failed scoring, it could be - notebook out of memory", "votes": null }, { "id": "2727818", "postDate": "04/02/2024 00:20:52", "content": "
\nLog says it ran suceeded but scoring is failed
![\"image\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F568032%2F34928b9cffefeceaa57b961592029719%2F2024-04-02%20%209.21.18.png?generation=1712017297304300&alt=media\")
\nOutput of the sample test set is exactly 10. You meant, there could be a problem in real test set (that is hidden)?
You are RIGHT. Thank you so much. I figured out that I didn't do the aggregate with 'other_1' file. I thought it has only 1 row for case_id, as it is in the training data set. It might have one or more rows for each unique case_id in actual test dataset! Thanks!
", "rawMarkdown": "You are RIGHT. Thank you so much. I figured out that I didn't do the aggregate with 'other_1' file. I thought it has only 1 row for case_id, as it is in the training data set. It might have one or more rows for each unique case_id in actual test dataset! Thanks!", "votes": null }, { "id": "2728260", "postDate": "04/02/2024 06:44:26", "content": "Try: submission['score'] = submission['score'].fillna(0)
", "rawMarkdown": "Try: submission['score'] = submission['score'].fillna(0)", "votes": null }, { "id": "2729017", "postDate": "04/02/2024 14:11:56", "content": "Glad to help! Can you give me some tip to improve score, or wether you have set up CV with correlation on LB? 🤝🤝🤝
", "rawMarkdown": "Glad to help! Can you give me some tip to improve score, or wether you have set up CV with correlation on LB? 🤝🤝🤝", "votes": null }, { "id": "2731962", "postDate": "04/03/2024 00:11:14", "content": "Sorry, I don't get the decent score right now with the ones I made on my own. .. around 0.374 only.
", "rawMarkdown": "Sorry, I don't get the decent score right now with the ones I made on my own. .. around 0.374 only.", "votes": null } ], "comments": [ { "id": 2726969, "author_name": "davidcanorosillo", "author_url": "", "post_date": "04/01/2024 14:06:42", "content": "Hey I had a similar bug driving me crazy a couple days ago. Basically I was generating multiple case id predictions because of bad aggregations. So make sure you are only outputting 10 rows of predictions (case id is always without duplicated).
\nGood luck!
\nOutput of the sample test set is exactly 10. You meant, there could be a problem in real test set (that is hidden)?
You are RIGHT. Thank you so much. I figured out that I didn't do the aggregate with 'other_1' file. I thought it has only 1 row for case_id, as it is in the training data set. It might have one or more rows for each unique case_id in actual test dataset! Thanks!
", "votes": null, "replies": [ { "id": 2729017, "author_name": "davidcanorosillo", "author_url": "", "post_date": "04/02/2024 14:11:56", "content": "Glad to help! Can you give me some tip to improve score, or wether you have set up CV with correlation on LB? 🤝🤝🤝
", "votes": null, "replies": [ { "id": 2731962, "author_name": "gilgarad", "author_url": "", "post_date": "04/03/2024 00:11:14", "content": "Sorry, I don't get the decent score right now with the ones I made on my own. .. around 0.374 only.
", "votes": null, "replies": [] } ] } ] } ] }, { "id": 2726985, "author_name": "eu1234", "author_url": "", "post_date": "04/01/2024 14:19:29", "content": "In submissions section check the error written under the version number of failed scoring, it could be - notebook out of memory
", "votes": null, "replies": [ { "id": 2727818, "author_name": "gilgarad", "author_url": "", "post_date": "04/02/2024 00:20:52", "content": "
\nLog says it ran suceeded but scoring is failed
Try: submission['score'] = submission['score'].fillna(0)
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2726530": "Hello, all.\n\nI've made my own notebook and run everything OK with the notebook itself.\nAnd I've submitted the notebook but I got the error of submission scoring error.\nI checked the log and everything was fine, and I can even see the predicted / created file of sample test set.\n\nI have no idea why this error occurs...\n\nDoes any of you experiences like me?", "2726969": "Hey I had a similar bug driving me crazy a couple days ago. Basically I was generating multiple case id predictions because of bad aggregations. So make sure you are only outputting 10 rows of predictions (case id is always without duplicated).\nGood luck!", "2726985": "In submissions section check the error written under the version number of failed scoring, it could be - notebook out of memory", "2727818": "\nLog says it ran suceeded but scoring is failed", "2727821": "\nOutput of the sample test set is exactly 10. You meant, there could be a problem in real test set (that is hidden)?", "2728249": "You are RIGHT. Thank you so much. I figured out that I didn't do the aggregate with 'other_1' file. I thought it has only 1 row for case_id, as it is in the training data set. It might have one or more rows for each unique case_id in actual test dataset! Thanks!", "2728260": "Try: submission['score'] = submission['score'].fillna(0)", "2729017": "Glad to help! Can you give me some tip to improve score, or wether you have set up CV with correlation on LB? 🤝🤝🤝", "2731962": "Sorry, I don't get the decent score right now with the ones I made on my own. .. around 0.374 only." }, "source": "meta" }