Hi all — I have a quick question regarding streaming behavior. I’m currently integrating BAML into a LangChain-compatible chat model (ChatBaml).

python
async def _astream()
  ...
  stream = self.b.stream.ChooseTool(baml_state, {"tb": tb}) # tb for @@dynamic usage in baml_src/
  async for partial in stream:
    ...

Right now partial.model_dump() is accumulated, e.g.:

- 1st emit: {"selected_tool": None}
- 2nd emit: {"selected_tool": {"name": None}}
- 3rd emit: {"selected_tool": {"name": "rep"}}
- 4th emit: {"selected_tool": {"name": "reply_to_user"}}

For LangChain streaming, I’d prefer delta-style updates (only the changed field/value per emit). Is there a configuration to make partial emit delta-style updates instead of accumulated state?

Hmm, could you indicate how you would want delta structured values to be represented?

In my case, since the

DynamicSchema

is already defined using

TypeBuilder

, I was expecting the delta to be a fully shaped dictionary with all fields present but set to None, for example:

python
{
    selected_tool: {
        name: None, # could be a delta or accumulated value; mainly need the full schema shape
        arguments: {
            a: None,
            b: None,
        },
    } or {
        name: None,
        arguments: {
            x: None,
            y: None,
        },
    } or {
        name: None,
        arguments: {
            role: "assistant", # constant
            content: None,
        },
    },
}

instead of:

python
{"selected_tool": None}

This would make it easier for me to compute delta-style updates on the LangChain side. That said, please feel free to ignore this if it doesn’t align with the intended design. Also, thanks for pointing out on_tick — that’s new to me, and I’ll take a look at the documentation.