There are many LLMs in the market. GPT4, Claude, Gemini, Mistral, etc. But which LLM suits your needs the best and how to choose it? LLM benchmarks are often confusing. As a user which benchmark score should you be concerned about while choosing the right LLM? Here I am curating a list of popular(not exhaustive) benchmarks and what they aim to measure. 1. SQuAD (Stanford Question Answering Dataset): Tests reading comprehension and answering questions based on given passages. 2. RACE: Evaluate understanding of passages from real-world scenarios through multiple-choice questions. 3. HellaSwag: Tests common sense by predicting what could happen next in a situation. 4. LAMBADA: Checks if a model can predict the right final word in a passage by understanding the context. 5. BoolQ: Yes/no questions to be answered based on short passages. Tests query understanding. 6. MultiRC: Requires answering multiple choice questions by reasoning over multiple passages. 7. ARC: Multiple-choice reasoning questions requiring gathering information from a science database. 8. CommonsenseQA: Tests commonsense reasoning through multiple-choice questions about everyday scenarios. 9. OpenBookQA: Science multiple-choice questions to be answered using an open book of facts. 10. GSM8K: Tests advanced math word problem solving needing multi-step reasoning. 11. CodeXGLUE: Evaluates code understanding and generation abilities across programming languages. 12. APPS: Assesses program understanding skills for Python code. 13. HumanEval: Tests coding abilities by having the model write functional programs. 14. HumanEval-X: An extension of HumanEval with more challenging coding tasks. 15. PIQA: Questions about diagrams and situations to test physical reasoning abilities. 16. VCR: Tests understanding of situations depicted in images through question-answering. 17. NLVR2: Determines if a statement accurately describes a set of images. 18. MMLU: Combines text with images/videos to test multimodal language and vision skills. So choose your LLM by looking at their score on your desirable benchmark.
I recommend this great blog post that provides guidance on using benchmarks to choose base and chat LLMs: https://osanseviero.github.io/hackerllama/blog/posts/llm_evals/
16. VCR sounds interesting. Any links/ references for this ?
MSCS @ Northeastern University | Prev - ML @Deloitte | Making GenAI systems run fast!
5moThis is a pretty good list. But unfortunately the relevancy of publically available benchmarks keeps changing because a lot of models on huggingface try to contaminate their training data with test samples to improve bench scores. They are important, nonetheless and just another metric that can be considered and not absolute! Often the best benchmark is testing the model out yourself (if you can, say, in a playground) and looking at the comments about it on r/LocalLLaMA.