BUG: Limit Azure OpenAI embeddings chunk size (#13425)

Hi! 
This short PR aims at:
* Fixing `OpenAIEmbeddings`' check on `chunk_size` when used with Azure
OpenAI (thus with openai < 1.0). Azure OpenAI embeddings support at most
16 chunks per batch, I believe we are supposed to take the min between
the passed value/default value and 16, not the max - which, I suppose,
was introduced by accident while refactoring the previous version of
this check from this other PR of mine: #10707
* Porting this fix to the newest class (`AzureOpenAIEmbeddings`) for
openai >= 1.0

This fixes #13539 (closed but the issue persists).  

@baskaryan @hwchase17

libs/langchain/langchain/embeddings/azure_openai.py

@@ -83,6 +83,10 @@ class AzureOpenAIEmbeddings(OpenAIEmbeddings):
         values["azure_ad_token"] = values["azure_ad_token"] or os.getenv(
             "AZURE_OPENAI_AD_TOKEN"
         )
+        # 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
+        values["chunk_size"] = min(values["chunk_size"], 16)
         try:
             import openai
 

libs/langchain/langchain/embeddings/openai.py

@@ -291,7 +291,7 @@ class OpenAIEmbeddings(BaseModel, Embeddings):
             # 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
-            values["chunk_size"] = max(values["chunk_size"], 16)
+            values["chunk_size"] = min(values["chunk_size"], 16)
         else:
             default_api_version = ""
         values["openai_api_version"] = get_from_dict_or_env(

libs/langchain/tests/integration_tests/embeddings/test_azure_openai.py

@@ -30,12 +30,25 @@ def test_azure_openai_embedding_documents_multiple() -> None:
     embedding = _get_embeddings(chunk_size=2)
     embedding.embedding_ctx_length = 8191
     output = embedding.embed_documents(documents)
+    assert embedding.chunk_size == 2
     assert len(output) == 3
     assert len(output[0]) == 1536
     assert len(output[1]) == 1536
     assert len(output[2]) == 1536
 
 
+def test_azure_openai_embedding_documents_chunk_size() -> None:
+    """Test openai embeddings."""
+    documents = ["foo bar"] * 20
+    embedding = _get_embeddings()
+    embedding.embedding_ctx_length = 8191
+    output = embedding.embed_documents(documents)
+    # Max 16 chunks per batch on Azure OpenAI embeddings
+    assert embedding.chunk_size == 16
+    assert len(output) == 20
+    assert all([len(out) == 1536 for out in output])
+
+
 @pytest.mark.asyncio
 async def test_azure_openai_embedding_documents_async_multiple() -> None:
     """Test openai embeddings."""