GitHub - arthur-ai/bench: A tool for evaluating LLMs
github.com
GitHub - arthur-ai/bench: A tool for evaluating LLMs
"My benchmark for large language models"
https://t.co/YZBuwpL0tl
Nice post but even more than the 100 tests specifically, the Github code looks excellent - full-featured test evaluation framework, easy to extend with further tests and run against many... See more
ANY
LLM of your choice, statistical methods, or NLP models that runs
locally on your machine
:
- G-Eval
- Summarization
- Answer Relevancy
- Faithfulness
- Contextual Recall
- Contextual Precision
- RAGAS
- Hallucination
- Toxicity
- Bias
- etc.