Can I specify a different model?

Yes, you can specify a different model for running evals.

Currently, we support the following OpenAI models:

• gpt-4-1106-preview
• gpt-4
• gpt-3.5-turbo

from athina.evals import LlmEvaluator
 
LlmEvaluator(model="gpt-4", grading_criteria=grading_criteria).run(response)

If you'd like to use a different model, contact us at hello@athina.ai.