jeff3071 commented on code in PR #71437:
URL: https://github.com/apache/airflow/pull/71437#discussion_r4109974976
##########
providers/common/ai/docs/connections/pydantic_ai.rst:
##########
@@ -54,14 +56,37 @@ Model
The model can also be overridden at the hook/operator level via the
``model_id`` parameter.
+Embedding Model
+ The embedding model identifier in ``provider:model`` format. This field
+ appears as a dedicated input in the connection form and stores its value in
+ ``extra["embed_model"]``.
+
+ Example: ``openai:text-embedding-3-small``
+
+ The embedding model and connection can also be overridden at the hook level
+ via the ``embed_model_id`` and ``embed_conn_id`` parameters.
+
+ When the LLM and embedding model use different providers, configure a
separate
+ ``embed_conn_id`` if the shared connection produces explicit configuration
for the
+ embedding provider. The providers can share a connection when no embedding
provider
+ configuration can be built from it; pydantic-ai then resolves the
embedding provider
+ independently, for example from environment variables, even if the LLM
uses configuration
+ from the connection. Equivalent OpenAI and Azure chat/response prefixes
can share their
+ provider's connection. Local ``sentence-transformers:`` embeddings can
also share the
+ LLM connection because they do not use provider credentials.
+
API Key (Password field)
- The API key for your LLM provider. Required for API-key-based providers
+ The API key for your model provider. Required for API-key-based providers
(OpenAI, Anthropic, Groq, Mistral). Leave empty for providers using
environment-based auth (Bedrock via ``AWS_PROFILE``, Vertex via
``GOOGLE_APPLICATION_CREDENTIALS``).
+ For Bedrock and Google models, provider-specific values from Extra are used
+ instead. A populated Password or Host is ignored for those model prefixes;
+ the hook emits a warning identifying the replacement Extra fields.
+
Review Comment:
Thanks for reviewing! Updated this part.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
