[ML] Make warnings from inference errors (#81735) #81761
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Consistently using exceptions for errors instead of WarningInferenceResults to simplify debugging/triaging # Conflicts: # x-pack/plugin/ml/qa/native-multi-node-tests/src/javaRestTest/java/org/elasticsearch/xpack/ml/integration/PyTorchModelIT.java # x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/nlp/FillMaskProcessor.java # x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/nlp/NerProcessor.java # x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/inference/nlp/FillMaskProcessorTests.java # x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/inference/nlp/NerProcessorTests.java # x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/inference/nlp/TextClassificationProcessorTests.java
davidkyle
added
:ml
Collaborator
|
Pinging @elastic/ml-core (Team:ML) |
| ) { | ||
| if (tokenization.getTokenizations().isEmpty() || tokenization.getTokenizations().get(0).getTokens().length == 0) { | ||
| return new WarningInferenceResults("No valid tokens for inference"); | ||
| throw new ElasticsearchStatusException("tokenization is empty", RestStatus.INTERNAL_SERVER_ERROR); |
Contributor
There was a problem hiding this comment.
Isn't this a bad request exception? If yes, we should reuse ExceptionsHelpers.badRequestException. Same stands for all the other exceptions in the PR that are marked as internal server error.
Member
Author
There was a problem hiding this comment.
The tokenisation is checked before the request is sent, for fill mask the error will be about a missing mask token. Here at results processing, if the tokenisation is not present something has gone badly wrong as we do not send the request if the tokenisation is not present.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Consistently using exceptions for errors instead of WarningInferenceResults. Using a mixture of the 2 complicates debugging and triaging as more both warnings and exceptions have to be checked.
Many of the uses of WarningInferenceResults related to validating the output of the native pytorch process. Those unexpected results are more accurately internal server errors not a warning.
Backport of #81735