21 lines
459 B
21 lines
459 B
from swarms.structs import Agent
|
|
from swarms.models.gpt4_vision_api import GPT4VisionAPI
|
|
from swarms.prompts.multi_modal_autonomous_instruction_prompt import (
|
|
MULTI_MODAL_AUTO_AGENT_SYSTEM_PROMPT_1,
|
|
)
|
|
|
|
|
|
llm = GPT4VisionAPI()
|
|
|
|
task = "What is the color of the object?"
|
|
img = "images/swarms.jpeg"
|
|
|
|
## Initialize the workflow
|
|
agent = Agent(
|
|
llm=llm,
|
|
sop=MULTI_MODAL_AUTO_AGENT_SYSTEM_PROMPT_1,
|
|
max_loops="auto",
|
|
)
|
|
|
|
agent.run(task=task, img=img)
|