Hierarchical process error: Tasks cannot be flagged with async_execution.

[ken4ward]

I am trying to create an app using CrewAI and it returns the error below. can anyone help out on what could be the issue?

This is Task file

# tasks.py
from crewai import Task
from crewai_apps.my_agents.ResearchAgent import data_analysis_agent, prediction_agent
from crewai_tools import DirectoryReadTool, FileReadTool

crypto_data_tool = DirectoryReadTool(directory='./data/COMPUSDT.csv')
file_tool = FileReadTool()

data_analysis_task = Task(
    description='Analyze cryptocurrency data to identify trends and patterns',
    expected_output='A detailed report of trends and patterns in the data',
    agent=data_analysis_agent,
    tools=[crypto_data_tool, file_tool],
    async_execution=True,
    callback=lambda output: print(f"Data Analysis Task completed!\nDescription: {output.description}\nOutput: {output.raw_output}")
)

prediction_task = Task(
    description='Predict future cryptocurrency prices based on analyzed data',
    expected_output='A report with predicted prices for the next month',
    agent=prediction_agent,
    tools=[file_tool],
    async_execution=True,
    callback=lambda output: print(f"Prediction Task completed!\nDescription: {output.description}\nOutput: {output.raw_output}")
)

This is the Agent file

# agents.py
import os
from crewai import Agent
from crewai_tools import SerperDevTool, DirectoryReadTool, FileReadTool

os.environ["OPENAI_API_KEY"] = "NA"

crypto_data_tool = DirectoryReadTool(directory='./data/COMPUSDT.csv')
file_tool = FileReadTool()
search_tool = SerperDevTool()

data_analysis_agent = Agent(
    role='Data Analyst',
    goal='Analyze cryptocurrency data to identify trends and patterns',
    backstory='A data analyst experienced in financial markets and cryptocurrency.',
    tools=[crypto_data_tool, file_tool],
    verbose=True
)

prediction_agent = Agent(
    role='Predictor',
    goal='Predict future cryptocurrency prices based on analyzed data',
    backstory='A machine learning expert with a focus on predictive modeling.',
    tools=[file_tool],
    verbose=True
)

# crew.py
import os

from crewai import Crew, Process
from dotenv import load_dotenv

from .my_agents.ResearchAgent import data_analysis_agent, prediction_agent
from .my_tasks.ResearchTasks import data_analysis_task, prediction_task
from langchain_community.chat_models.ollama import ChatOllama
from langchain_community.llms.ollama import Ollama

load_dotenv()

# memory_system = MemorySystem(
#     short_term=True,
#     long_term=True,
#     entity=True,
#     contextual=True,
#     embedder={"provider": "openai", "config": {"model": 'text-embedding-3-small'}}
# )

llm = Ollama(
    model=os.environ['OPENAI_MODEL_NAME'],
    base_url=os.environ['OPENAI_API_BASE']
)

crew = Crew(
    agents=[data_analysis_agent, prediction_agent],
    tasks=[data_analysis_task, prediction_task],
    process=Process.hierarchical,
    manager_llm=ChatOllama(model="llama3.1"),
    memory_system=True,
    verbose=True,
    cache=True,
    full_output=True,
    output_log_file='crew_output.log',
    llm=llm
)

This is the error

Traceback (most recent call last):
  File "/Users/christdoes/Documents/projects/AILLM/CrewAI/projects/crewai-apps/main.py", line 2, in <module>
    from crewai_apps.ResearchCrew import crew
  File "/Users/christdoes/Documents/projects/AILLM/CrewAI/projects/crewai-apps/crewai_apps/ResearchCrew.py", line 27, in <module>
    crew = Crew(
           ^^^^^
  File "/Users/christdoes/anaconda3/lib/python3.12/site-packages/pydantic/main.py", line 193, in __init__
    self.__pydantic_validator__.validate_python(data, self_instance=self)

I'll really appreciate if anyonr can help look into it.

[pythonbyte]

Hey @ken4ward, thanks for adding the issue please follow the Issue template with [BUG] on the beginning of the title and the following information

Steps to Reproduce Provide a step-by-step process to reproduce the behavior:

Screenshots/Code snippets If applicable, add screenshots or code snippets to help explain your problem.

Environment Details:

• Operating System: [e.g., Ubuntu 20.04, macOS Catalina, Windows 10]
• Python Version: [e.g., 3.8, 3.9, 3.10]
• crewAI Version: [e.g., 0.30.11]
• crewAI Tools Version: [e.g., 0.2.6]

[lorenzejay]

I believe the next release will fix this issue. To use async_execution on your tasks, use sequential process for now or build from our main