grouse - 0.4.2

Fixed

• Fix completion cost computation to support more models with Litellm like vertex_ai models

- Python
Published by antonioloison over 1 year ago

grouse - 0.4.1

Fixed

• Fix API Connection Errors with Semaphore

- Python
Published by antonioloison over 1 year ago

grouse - 0.4.0

Added

• Added o1 support
• Added meta evaluation pipeline to Python package

Fixed

• Change jinja template extensions

- Python
Published by antonioloison over 1 year ago

grouse - 0.3.1

Fixed

• Fixed prompts newlines
• Removed latest structured generation because faithfulness and usefulness dtos did not support it
• Fixed dataset loading in plot function

- Python
Published by antonioloison over 1 year ago

grouse - 0.3.0

Added

• Add flag to use the training dataset in meta evaluation.

- Python
Published by antonioloison over 1 year ago

grouse - 0.2.1

Fixed

• Add missing model register file.

- Python
Published by antonioloison over 1 year ago

grouse - 0.2.0

Added

• Register Fireworks Llama 3.1 8b and 70b prices with litellm to better support these models as evaluators.

Changed

• Remove instructor package to better understand what is really sent to the LLM. All the LLM generations are simply done using litellm.acompletion.
• Removed black from justfile and dev dependencies as ruff plays the same role.

- Python
Published by antonioloison over 1 year ago

grouse - 0.1.0

Added

• Created GroundedQAEvaluator that evaluates four metrics per sample: answer relevancy, completeness, faithfulness, usefulness, negative rejection and positive acceptance.
• Created MetaEvaluator to evaluate evaluators on GroUSE unit tests.
• Added plot sub-command to show the matrices

- Python
Published by antonioloison over 1 year ago