Regarding Language Studio's auto-labeling problem

yamada kyohei 0

When I run auto-labeling using a model that has already been created in Language Studio, the auto-labeling succeeds, but no labeling is done and the consumed quota ends up being 0. Do you know what the cause is? I couldn't find the cause because it doesn't happen in all jobs, but only occasionally.

romungi-MSFT 44,661 Reputation points Microsoft Employee

2024-01-22T06:40:38.29+00:00

@yamada kyohei Did you get a chance to review above comment? Thanks!!
Nathan Gonzalez 0 Reputation points

2024-03-19T14:01:07.88+00:00

I have this exact problem. I select an existing custom ner model, several documents, and then the job succeeds with 0 quota used and no documents being labeled. Sometimes it works, but even then not for all documents selected, just a seemingly random subset.

1 answer

romungi-MSFT 44,661 Reputation points Microsoft Employee

2024-01-18T07:00:43.5+00:00

@yamada kyohei Do you mean labeling using GPT or labeling using a model that is already trained? While using GPT the token limits could be used, and the quota is consumed. For already trained model in language studio are you seeing any token quota being consumed? If yes, could you please add more details, like a screen shot of the page or portal?
Please sign in to rate this answer.

0 comments No comments
Sign in to comment

Use comments to ask for clarification, additional information, or improvements to the question.

Share via

Regarding Language Studio's auto-labeling problem

1 answer

Your answer