measurement
Time to First Token (TTFT) is a performance metric for LLMs, often used in service level agreements (SLAs) where organizations target a 95th percentile response time of under 2 seconds for user queries.
Authors
Sources
- LLM Observability: How to Monitor AI When It Thinks in Tokens | TTMS ttms.com via serper
Referenced by nodes (1)
- Large Language Models concept