Kaggleのcodeコンペのサブミット時のエラーについてのメモ

kaggleのcodeコンペで、サブミットした時に出るerrorについてです。

• Notebook Timeout: Your submission notebook exceeded the allowed runtime. Review the competition's Code Requirements page for the time limits. Note that the hidden dataset can be larger/smaller/different than the public dataset.
• Notebook Threw Exception: While rerunning your code, your notebook hit an unhandled error. Note that the hidden dataset can be larger/smaller/different than the public dataset. Due to the way Kaggle handles notebook runtime exceptions, this error class is currently only possible with scripts (not notebooks).
• Notebook Exceeded Allowed Compute: This indicates you have violated a code requirement constraint during the rerun. This includes limitations in the execution environment, for example requesting more RAM or disk than available, or competition constraints, such as input data source type or size limits.
• Submission CSV Not Found: Your notebook did not output the expected submission file (often submission.csv). The rerun of your notebook appears to have completed, but when we looked for your submission file, it wasn't there. This means it's possible to have bugs/errors upstream that halted execution and prevented the file from being written. Attempting to read a non-existent directory/file from a dataset is a common reason execution is halted (causing either Submission CSV Not Found or Notebook Threw Exception).
• Submission Scoring Error: Your notebook generated a submission file with incorrect format. Some examples causing this are: wrong number of rows or columns, empty values, an incorrect data type for a value, or invalid submission values from what is expected.
• Kaggle Error: A rare system error. Please try resubmitting to resolve the error and contact Kaggle Support if it persists.

Code Competitions - Errors & Debugging Tips | Kaggle より

毎回エラーがどこに書いてあるのかわからなくなるのでメモっておきました。

【2022.9.30 追記】

One thing to keep in mind: I'm pretty sure that running out of RAM gives you "Notebook Exceeded Allowed Compute", but running out of GPU RAM gives you "Notebook Threw Exception," which confused me at first.

So, I always first make sure to reduce batch sizes.

— Mark Tenenholtz (@marktenenholtz) September 29, 2022

"Notebook Exceeded Allowed Compute"はCPUのメモリエラー
"Notebook Threw Exception"はGPUのメモリエラー
の可能性という素晴らしい知見をもらった https://t.co/2XJNgFj7Ps

— カレーちゃん🍮 (@currypurin) September 29, 2022

Notebook Exceeded Allowed ComputeとNotebook Threw Exceptionでこの違いがあるとのこと。