[DOC] Add NONE pooling mode to text embedding model documentation

[aneesh-db]

What do you want to do?

• Request a change to existing documentation
• Add new documentation
• Report a technical problem with the documentation
• Other

Tell us about your request.

The pooling_mode parameter in the Register Model API documentation (_ml-commons-plugin/api/model-apis/register-model.md) currently lists supported values as mean, mean_sqrt_len, max, weightedmean, and cls.

A new none pooling mode is being added in ml-commons (opensearch-project/ml-commons#4710) to fix a bug where MEAN pooling was applied by default even for models that already provide pre-pooled sentence_embedding output. This caused embeddings to not match Python inference output.

The documentation should be updated to:

1. Add none to the list of supported pooling_mode values in the model_config object table
2. Describe that none uses pre-pooled output from the model directly without additional pooling computation, suitable for models that already provide pooled embeddings (e.g., sentence_embedding or pooler_output)

Version: 3.4

What other resources are available?

• ml-commons issue: [BUG] MEAN pooling applied by default when no pooling should be used for pre-pooled model outputs ml-commons#4708
• ml-commons PR: Add NONE pooling mode to support pre-pooled model outputs ml-commons#4710

[kolchfa-aws]

@aneesh-db Is this also going into 3.6?

[aneesh-db]

@kolchfa-aws PR is merged now.

[kolchfa-aws]

@aneesh-db Thanks! Would you be willing to submit a documentation PR for this?

[aneesh-db]

sure @kolchfa-aws , you can assign this to me