Update DeepSeekProvider to comply with DeepSeek v3.2 requirements for sending thinking parts

Author: ZeroAurora

pydantic_ai_slim/pydantic_ai/providers/deepseek.py

@@ -47,9 +47,8 @@ def model_profile(self, model_name: str) -> ModelProfile | None:
         return OpenAIModelProfile(
             json_schema_transformer=OpenAIJsonSchemaTransformer,
             openai_chat_custom_reasoning_field='reasoning_content',
-            # DeepSeek recommends against sending back unchanged reasoning parts in requests.
-            # The following is for compatibility with existing behavior. May want to change later.
-            openai_chat_send_back_thinking_parts='thinking_tags',
+            # Starting from DeepSeek v3.2, DeepSeek requires sending thinking parts for optimal agentic performance.
+            openai_chat_send_back_thinking_parts='custom_field',
         ).update(profile)
 
     @overload