trinityml 0.0.10

A Python SDK for recording the LLM Experiment Execution

Installation

To install the Trinity package, run the following command:

pip install trinityml

Setup the Environment Variable

export TRINITY_SECRET_KEY="your_secret_key"
export TRINITY_PUBLIC_KEY="your_public_key"
export TRINITY_HOST="http://your_trinity_host"

Record the Experiment

To use the record function, ensure that you have properly set up your environment and then call the function as shown below:

# Import the Trinity package
import trinityml

# Call the record() function
trinityml.record(
    experiment_name=experiment_name,
    questions=questions,
    contexts=contexts,
    generated_prompts=generated_prompts,
    answers=answers,
    ground_truths=ground_truths,
    model_temperature=model_temperature,
    top_k=top_k,
    model_name=model_name,
    prompt_template=prompt_template
)