Recent submissions

Comment · 3 years ago
say using A10G ~$1.2/hr and with full utilization on vllm 112 reqs/min => per req ~$0.00018 versus gpt-3.5 turbo $0.002 per 1k token
Comment · 3 years ago
the cost would be depending on GPU type/serving system/traffic pattern. check out some throughput comparison from vllm's blog post https://vllm.ai/ if you serve 7B on cost-optimized GPUs (A10G/L4) and keep it busy, it can be a lot cheaper than gpt3.5 turbo. tho it's not a fair comparison as 3.5's quality is still far better.
Comment · 3 years ago
Comment · 3 years ago
Surprised to learn StableLM is worse than plain LLaMA. link to their leaderboard: leaderboard.lmsys.org
Comment · 3 years ago
Comment · 3 years ago
Comment · 3 years ago
just tried it worked okay for me tho.

> Can you name all of the soccer players whose names contain a Z

> Here are all the soccer players whose names contain a "Z" that I could find:

Zlatan Ibrahimovic Zinedine Zidane Zlatko Zahovic Zeljko Kalac Zoran Banovic Zlatko Arsovic Zoran Savic Zoran Jovanovic Zoran Djurdjic Zoran Radosavljevic Zoran Zekic Zoran Mirkovic Zoran Sretenovic Zoran Sosic Zoran Luburic Zoran Vulin Zlatan Ljubijankic Zoran Lekovic Zoran Culibrk Zoran Bojovic Please note that this list may not be exhaustive, and there may be other soccer players with names containing "Z" that are not included.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton