Update doc per feedback (elastic#137970)

Author: kderusso

docs/reference/elasticsearch/mapping-reference/semantic-text.md

@@ -280,8 +280,7 @@ endpoint associated with `inference_id`.
 If chunking settings are updated, they will not be applied to existing documents
 until they are reindexed.
 To completely disable chunking, use the `none` chunking strategy.
-
-Settings for chunking text into smaller passages for scoring and reranking. Defaults to the optimal chunking settings for [Elastic Rerank](docs-content:///explore-analyze/machine-learning/nlp/ml-nlp-rerank.md). Refer to the [Inference API documentation](https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-put#operation-inference-put-body-application-json-chunking_settings) for valid values for `chunking_settings`. 
+Defaults to the optimal chunking settings for [Elastic Rerank](docs-content:///explore-analyze/machine-learning/nlp/ml-nlp-rerank.md). Refer to the [Inference API documentation](https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-put#operation-inference-put-body-application-json-chunking_settings) for valid values for `chunking_settings`. 
 
 ::::{warning}
 When using the `none` chunking strategy, if the input exceeds the maximum token