Add config option to select fp16 or quantized jina vision model (#14270)

* Add config option to select fp16 or quantized jina vision model

* requires_fp16 for text and large models only

* fix model type check

* fix cpu

* pass model size

docs/docs/configuration/reference.md

@@ -520,6 +520,8 @@ semantic_search:
   reindex: False
   # Optional: Set device used to run embeddings, options are AUTO, CPU, GPU. (default: shown below)
   device: "AUTO"
+  # Optional: Set the model size used for embeddings. (default: shown below)
+  model_size: "small"
 
 # Optional: Configuration for AI generated tracked object descriptions
 # NOTE: Semantic Search must be enabled for this to do anything.