Turbo #LoRA just got better! No need to retrain your LoRA, just add Turbo to existing LoRAs, boosting throughput 2x with no quality loss. We’re now hitting 365 tokens/sec on top-tier hardware! Catch the full webinar replay: https://
youtube.com/watch?v=oWxt3U
YHSUU
… #AI #ML #llm

→ View original post on X — @predibase,