The step decorator will log steps based on the decorated function. By default, the arguments of the function will be used as the input of the step and the return value will be used as the output. Under the hood, the step decorator is using the cl.Step class. Parameters
The name of the step. Default to the name of the decorated function.
The type of the step, useful for monitoring and debugging.
show_input
Union[bool, str]
default:false
By default only the output of the step is shown. Set this to True to also show the input. You can also set this to a language like json or python to syntax highlight the input.
Access the Current step
You can access the current step object using cl.context.current_step and override values. import chainlit as cl @cl.step async def my_step(): current_step = cl.context.current_step # Override the input of the step current_step.input = "My custom input" # Override the output of the step current_step.output = "My custom output"
Stream the Output
from openai import AsyncOpenAI import chainlit as cl client = AsyncOpenAI(api_key="YOUR_API_KEY") @cl.step(type="llm") async def gpt4(): settings = { "model": "gpt-4", "temperature": 0, } stream = await client.chat.completions.create( messages=message_history, stream=True, **settings ) current_step = cl.context.current_step async for part in stream: delta = part.choices[0].delta if delta.content: # Stream the output of the step await current_step.stream_token(delta.content)
Nest Steps
If another step decorated function is called inside the decorated function, the child step will be nested under the parent step. import chainlit as cl @cl.step async def parent_step(): await child_step() return "Parent step output" @cl.step async def child_step(): return "Child step output" @cl.on_chat_start async def main(): await parent_step()
