fix(embeddings): number of texts in Azure OpenAIEmbeddings batch (lan…

…gchain-ai#10707)

This PR addresses the limitation of Azure OpenAI embeddings, which can
handle at maximum 16 texts in a batch. This can be solved setting
`chunk_size=16`. However, I'd love to have this automated, not to force
the user to figure where the issue comes from and how to solve it.

Closes langchain-ai#4575. 

@baskaryan

---------

Co-authored-by: Harrison Chase <[email protected]>

libs/langchain/langchain/embeddings/openai.py

@@ -231,7 +231,7 @@ def build_extra(cls, values: Dict[str, Any]) -> Dict[str, Any]:
         values["model_kwargs"] = extra
         return values
 
-    @root_validator()
+    @root_validator(pre=True)
     def validate_environment(cls, values: Dict) -> Dict:
         """Validate that api key and python package exists in environment."""
         values["openai_api_key"] = get_from_dict_or_env(
@@ -257,8 +257,13 @@ def validate_environment(cls, values: Dict) -> Dict:
         )
         if values["openai_api_type"] in ("azure", "azure_ad", "azuread"):
             default_api_version = "2022-12-01"
+            # Azure OpenAI embedding models allow a maximum of 16 texts
+            # at a time in each batch
+            # See: https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#embeddings
+            default_chunk_size = 16
         else:
             default_api_version = ""
+            default_chunk_size = 1000
         values["openai_api_version"] = get_from_dict_or_env(
             values,
             "openai_api_version",
@@ -271,6 +276,8 @@ def validate_environment(cls, values: Dict) -> Dict:
             "OPENAI_ORGANIZATION",
             default="",
         )
+        if "chunk_size" not in values:
+            values["chunk_size"] = default_chunk_size
         try:
             import openai