6.4 KiB

Raw Blame History

Local run with Docker or docker-compose

Environment Configuration with .env

This project uses a .env file to manage environment-specific configurations for local development, especially when using Docker Compose. This allows you to customize settings without modifying version-controlled files like docker-compose.yml.

Setup

Create your local .env file: Copy the example configuration to a new .env file:
```
cp .env.example .env
```
Customize your .env file: Open the .env file in a text editor and modify the variables as needed for your local setup. For example, you might want to change LLM models or API keys.

How it Works with Docker Compose

When you run docker-compose up or docker-compose run, Docker Compose automatically looks for a .env file in the project root directory (where docker-compose.yml is located) and loads the environment variables defined in it. These variables are then passed into the container environment for the app service.

The .env file itself is ignored by Git (as specified in .gitignore), so your local configurations will not be committed to the repository.

For streamlined experience, it is recommended to use docker compose as it simplifies experience eg handles mounting of directories for caching. Skip to [Docker Compose](#Using Docker Compose) section for this.

Running with Docker

This project supports running within a Docker container, which ensures a consistent environment for development and testing.

Prerequisites

Docker installed and running on your system.

Build the Docker Image

Navigate to the root directory of the project (where the Dockerfile is located) and run:

docker build -t tradingagents .

Test local ollama setup

To test ollama connectivity and local model:

docker run -it --env-file .env tradingagents python test_ollama_connection.py

for picking environment settings from .env file. You can pass values directly using:

docker run -it \
  -e LLM_PROVIDER="ollama" \
  -e LLM_BACKEND_URL="http://localhost:11434/v1" \
  -e LLM_DEEP_THINK_MODEL="qwen3:0.6b" \
  -e LLM_EMBEDDING_MODEL="nomic-embed-text"\ 
  tradingagents \
  python test_ollama_connection.py

For prevent re-downloading of Ollama models, mount folder from your host and run as

docker run -it \
  -e LLM_PROVIDER="ollama" \
  -e LLM_BACKEND_URL="http://localhost:11434/v1" \
  -e LLM_DEEP_THINK_MODEL="qwen3:0.6b" \
  -e LLM_EMBEDDING_MODEL="nomic-embed-text"\ 
  -v ./ollama_cache:/app/.ollma \
  tradingagents \
  python test_ollama_connection.py

Notes on Ollama for Local Docker: When LLM_PROVIDER is set to ollama the ollama server is automatically started in the docker conatiner. The LLM_BACKEND_URL is set to http://localhost:11434/v1. This assumes you have Ollama running on your host machine and accessible at port 11434. '/v1' is added to url at the end for OpenAI api compatibility.

Run the Main Application

To run the main.py script:

docker run -it --env-file .env tradingagents python -m main

  -e LLM_PROVIDER="ollama" \
  -e LLM_BACKEND_URL="http://localhost:11434/v1" \
  -e LLM_DEEP_THINK_MODEL="qwen3:0.6b" \
  -e LLM_QUICK_THINK_MODEL="qwen3:0.6b" \
  -e LLM_EMBEDDING_MODEL="nomic-embed-text"\ 
  -e MAX_DEBATE_ROUNDS="1" \
  -e ONLINE_TOOLS="False" \
  -v ./ollama_cache:/app/.ollama \
  tradingagents python -m main

Adjust environment variables as needed for your local setup.

Run the TradingAgents CLI

To run the cli interface (default in the container)

docker run -it --env-file .env tradingagents

or or

  -e LLM_PROVIDER="ollama" \
  -e LLM_BACKEND_URL="http://localhost:11434/v1" \
  -v ./ollama_cache:/app/.ollma \
  tradingagents python -m cli.main

Adjust environment variables as needed for your local setup.

Using Docker Compose

For a more streamlined local development experience, it is recommended to use Docker Compose. The docker-compose.yml file in the project root is configured to use the existing Dockerfile.

Build and Run Tests:

The default command in docker-compose.yml is set to run the test suite.

docker-compose up --build

This command will build the image (if it's not already built or if changes are detected) and then run the pytest tests/test_main.py command. The --rm flag is implicitly handled by up when the process finishes, or you can run:

docker-compose run --it app # This will use the default command from docker-compose.yml

If you want to explicitly run the tests:

docker-compose run --rm app python test_ollama_connection.py

Run the Main Application:

To run the main.py script, you can override the default command:

docker-compose run --rm app python -m main

Or, you can modify the command in docker-compose.yml if you primarily want docker-compose up to run the main application.

Run the TradingAgents CLI Application:

To run the cli/main.py script, you can override the default command:

docker-compose run --it app python -m cli.main

Environment Variables: The necessary environment variables (like LLM_PROVIDER, LLM_BACKEND_URL, model names, etc.) are configured in the docker-compose.yml for the app service. Ollama is started by the entrypoint script within the same container when LLM_PROVIDER is set to ollama, and LLM_BACKEND_URL is set to http://localhost:11434/v1.

When using enviroment file for docker command, please do not put extra quotes around the values, or any extra comments at the end otherwise docker will not pick the values.

Live Code Reloading: The current directory is mounted as a volume into the container at /app. This means changes you make to your local code will be reflected inside the container, which is useful for development. You might need to rebuild the image with docker-compose build or docker-compose up --build if you change dependencies in requirements.txt or modify the Dockerfile itself.

Ollama Model Caching: To prevent re-downloading Ollama models, docker-compose.yml now mounts ./.ollama on your host to /app/.ollama in the container. Models pulled by Ollama will be stored in ./.ollama/models locally and persist across runs. Ensure this directory is in your .gitignore. If Docker has permission issues creating this directory, you might need to create it manually (mkdir .ollama).

6.4 KiB Raw Blame History