Good checklist on Evaluating LLM Models
- LLM Type - Open source / Proprietary / Datasets
- Deployment Options - Private cloud or API model
- Infra needs - Self-hosting
- Retrieval Augmentation - Support for RAG
- Scalability - Performance of large RAG datasets
- Hallucinations - Handling Hallucinations
- Benchmark compared to other competitive models
- Legal Compliance - IP / ownership of prompts / PII
- Output Compliance - Bias / Toxicity
- Output filtering / Content filters
Keep Exploring!!!
No comments:
Post a Comment