-
Notifications
You must be signed in to change notification settings - Fork 0
/
6_code_executor_chart.py
58 lines (47 loc) · 2.13 KB
/
6_code_executor_chart.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
# https://microsoft.github.io/autogen/docs/tutorial/code-executors/
import os
from dotenv import load_dotenv
import tempfile
import datetime
# For 'import autogen' to work, had to 'pip install packaging'
from autogen import AssistantAgent, UserProxyAgent
from autogen.coding import DockerCommandLineCodeExecutor
load_dotenv()
# For some reason I need proxyman running and OPENAI_BASE_URL set to http://localhost:10800/v1 for this to work 🤷♂️
llm_config = {
'model': 'gpt-4o',
'api_key': os.getenv('OPENAI_API_KEY'),
'base_url': os.getenv('OPENAI_BASE_URL') if os.getenv('OPENAI_BASE_URL') else None,
}
# Create a temporary directory to store the code files.
# On macOS, the temporary directory gets created here and is automatically deleted when this script finishes.
# /private/var/folders/pz/{random_string}/T/tmp{random_string}
temp_dir = tempfile.TemporaryDirectory()
# Create a Docker command line code executor.
executor = DockerCommandLineCodeExecutor(
# Execute code using the given docker image name.
image="python:3.12-slim",
timeout=10, # Timeout for each code execution in seconds.
# Use the temporary directory to store the code files.
work_dir=temp_dir.name,
)
code_executor_agent = UserProxyAgent(
"code_executor_agent",
# Use the docker command line code executor.
code_execution_config={"executor": executor},
)
code_writer_agent = AssistantAgent(
"code_writer_agent",
llm_config=llm_config,
)
today = datetime.datetime.now().strftime("%Y-%m-%d")
chat_result = code_executor_agent.initiate_chat(
code_writer_agent,
message=f"Today is {today}. Write Python code to plot TSLA's and META's "
"stock price gains YTD, and save the plot to a file named 'stock_gains.png'.",
)
# Before the code is terminated (but after the chart is generated), we can view the generated file by:
# - Check where the mounted temporary directory is located: docker inspect -f '{{json .Mounts}}' {container_id}
# - Copy that folder to somewhere persistent: cp -r /path/to/temporary/directory ~/Downloads/
# When the code executor is no longer used, stop it to release the resources.
# executor.stop()