v0.2.4

v0.2.4

v0.2.3

v0.2.3

v0.2.1

v0.2.1

v0.2.0

v0.2.0 is out now, with the main improvement being the addition of streaming support!

Now, if you have a workflow that writes to the event stream like:

class ProgressEvent(Event):
    progress: str


# create a dummy workflow
class MyWorkflow(Workflow):
    @step()
    async def run_step(self, ctx: Context, ev: StartEvent) -> StopEvent:
        # Your workflow logic here
        arg1 = str(ev.get("arg1", ""))
        result = arg1 + "_result"

        # stream events as steps run
        ctx.write_event_to_stream(
            ProgressEvent(progress="I am doing something!")
        )

        return StopEvent(result=result)

You can stream the events using the client

# create a session
session = client.create_session()

# kick off run
task_id = session.run_nowait("streaming_workflow", arg1="hello_world")

# stream events -- the will yield a dict representing each event
for event in session.get_task_result_stream(task_id):
    print(event)

# get final result
result = session.get_task_result(task_id)
print(result)
# prints 'hello_world_result'

v0.1.3

v0.1.3

v0.1.2

v0.1.2

v0.1.1

v0.1.1

v0.1.0

llama_agents -> llama_deploy

v0.1.0 is here! This is a huge refactor, which also renames llama_agents to llama_deploy !

Now, llama_deploy is the place to go to deploy and scale agentic workflows that you built with llama_index.

We have extensive documentation in the updated readme, and thorough examples, so check it out!

We've also added official docs pages and API reference to the llama_index documentation for llama_deploy !

v0.0.14

v0.0.14

v0.0.13

v0.0.13