Docker and local code is not executing [Ollama]

[fahadshery]

Hi,

I am using the following:

PRETTY_NAME="Ubuntu 24.04.1 LTS"
NAME="Ubuntu"
VERSION_ID="24.04"
VERSION="24.04.1 LTS (Noble Numbat)"
VERSION_CODENAME=noble
ID=ubuntu
ID_LIKE=debian
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
UBUNTU_CODENAME=noble
LOGO=ubuntu-logo

I have docker installed:

docker --version
Docker version 27.3.1, build ce12230

I have the following installed:

pip freeze
annotated-types==0.7.0
anyio==4.6.0
certifi==2024.8.30
charset-normalizer==3.4.0
diskcache==5.6.3
distro==1.9.0
docker==7.1.0
FLAML==2.3.1
h11==0.14.0
httpcore==1.0.6
httpx==0.27.2
idna==3.10
jiter==0.6.1
numpy==1.26.4
openai==1.51.2
packaging==24.1
pyautogen==0.3.1
pydantic==2.9.2
pydantic_core==2.23.4
python-dotenv==1.0.1
regex==2024.9.11
requests==2.32.3
sniffio==1.3.1
termcolor==2.5.0
tiktoken==0.8.0
tqdm==4.66.5
typing_extensions==4.12.2
urllib3==2.2.3

This is my example script:

main.py
import os
import autogen
from autogen import AssistantAgent, UserProxyAgent

config_list = [
    {
    'model': "llama3.2:3b",
    'base_url': "https://ollama.local.example.com/v1",
    'api_key': "ollama",
  }
]

llm_config = {
    "config_list": config_list,
    "cache_seed": None,
    "price": [0,0],
}

# llm_config = {"model": "gpt-4", "api_key": os.environ["OPENAI_API_KEY"]}
# assistant = AssistantAgent("assistant", llm_config=llm_config)

with autogen.coding.DockerCommandLineCodeExecutor(work_dir="coding") as code_executor:
    assistant = AssistantAgent("assistant", llm_config=llm_config)
    user_proxy = UserProxyAgent(
        "user_proxy", code_execution_config={"executor": code_executor}
    )

    # Start the chat
    user_proxy.initiate_chat(
        assistant,
        message="Plot a chart of NVDA and TESLA stock price change YTD. Save the plot to a file called plot.png",
    )

but I don't see any output in the coding dir.
It just produces this output:

python main.py
flaml.automl is not available. Please install flaml[automl] to enable AutoML functionalities.
user_proxy (to assistant):

Plot a chart of NVDA and TESLA stock price change YTD. Save the plot to a file called plot.png

--------------------------------------------------------------------------------


# Calculate closing price change YTD
nvda_change = (nvda_data['Close'].iloc[-1] - nvda_data['Close'].iloc[0]) / nvda_data['Open'].iloc[0]
tesla_change = (tesla_data['Close'].iloc[-1] - tesla_data['Close'].iloc[0]) / tesla_data['Open'].iloc[0]

# Plot bar chart using matplotlib
plt.figure(figsize=(10, 6))
plt.bar(['NVDA', 'TSLA'], [nvda_change, tesla_change])
plt.xlabel('Stock')
plt.ylabel('YTD Change (%)')
plt.title('YTD Price Change of NVDA and TSLA')
plt.tight_layout()
plt.savefig('plot.png')

print("Plot saved to plot.png")
TERMINATE

Note: This script requires the yfinance library to download historical stock data. It also assumes that you have matplotlib installed, which is used for plotting. You can install these libraries using pip with the command pip install yfinance matplotlib.

Also, please note that this code saves a bar chart of YTD price change instead of the exact chart requested (which would require more complex plotting and possibly interactive features).

---

Replying as user_proxy. Provide feedback to assistant. Press enter to skip and use auto-reply, or type 'exit' to end the conversation:

what am I not doing?
I also tried installing:

pip install autogen-agentchat~=0.2

the result is the same...it just doesn't run code `locally` nor in `docker`

[ekzhu]

It looks like the local model didn't format the code in markdown blocks?

The code executor looks for markdown blocks for code