-
Notifications
You must be signed in to change notification settings - Fork 1
/
local_mistral.py
83 lines (74 loc) · 2.57 KB
/
local_mistral.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
import os
from crewai import Agent, Task, Crew, Process
from tools.custom_tools import CustomTools
from langchain.llms import Ollama
ollama_llm = Ollama(model="mistral")
from langchain.tools import DuckDuckGoSearchRun
search_tool = DuckDuckGoSearchRun()
# Define your agents with roles and goals
researcher = Agent(
role='Research Analyst',
goal='Research and create reports',
backstory="""You work at a leading research team.
Your expertise lies in research and find useful information.""",
verbose=True,
allow_delegation=False,
tools=[search_tool],
llm=ollama_llm,
)
note_taker = Agent(
role='Note Taker',
goal='Save content as note and store them as markdown on local device',
backstory="""You structure and format the given content to markdown
and create a note to be stored on disk.
You MUST use the TOOL provided to store the note to Obsidian.
If the content has a list use a markdown table to organize the text.
If the content starts or ends with ``` remove the ```.
Make sure the text is saved on disk.
When the note is saved successful return "Note is saved.".
""",
verbose=True,
allow_delegation=False,
tools=[CustomTools.store_note_to_obsidian],
llm=ollama_llm,
)
editor = Agent(
role='Content Editor',
goal='Summerize content and craft brief and useful notes',
backstory="""You are a renowned Content Editor, known for
sharp bried useful content.
You break complex topics into brief useful notes.""",
verbose=True,
allow_delegation=True,
llm=ollama_llm,
)
# Create tasks for your agents
task1 = Task(
description="""Conduct a brief research about School Holidays in Hamburg in 2024.
Define the start and end and duration.
Your final answer MUST be a brief report""",
agent=researcher
)
task2 = Task(
description="""Using the information provided, create a brief note.
If there is a listing in the content use markdown table to structure it.
Your final answer MUST be a note in markdown.""",
agent=editor
)
task3 = Task(
description="""Using the content provided, save the content as a note.
You must save the contetnt as a note to the Obsidian note taking App.
It is bvery important to store the content as a note to be used later.
Your final answer MUST be the information is the note is stored or not.""",
agent=note_taker
)
# Instantiate your crew with a sequential process
crew = Crew(
agents=[researcher, editor, note_taker],
tasks=[task1, task2, task3],
verbose=2, # You can set it to 1 or 2 to different logging levels
)
# Get your crew to work!
result = crew.kickoff()
print("######################")
print(result)