llm-experiments

random experiments with llm

usage

you need to run ollama serve to use deepseek and llama.

place a .env file at the root of the repository as shown in the example below:

# openai
OPENAI_API_KEY=***
# google for gemini
GOOGLE_API_KEY=***
# serper
SERPER_API_KEY=***
# tavily
TAVILY_API_KEY=***
# slack
SLACK_USER_TOKEN=***
# langsmith (optional)
LANGSMITH_TRACING=***
LANGSMITH_ENDPOINT=***
LANGSMITH_API_KEY=***
LANGSMITH_PROJECT=***

then run:

make install && source .env && source .venv/bin/activate

to start working with the swe team, run:

python -m llm_experiments --agent swe

if you just want a simple agent, run:

python -m llm_experiments

additionally, you can work with a more specialized agent by running:

python -m llm_experiments --agent slack

to see all available agent options, run:

python -m llm_experiments --help