SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second per-token latency

Re:SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second ... by whimsicalism on Sunday May 05, 2024 @02:59AM (#40262842)

I am curious whether this is true - OAI at least has the reputation in the industry of caring the least about safety of the major labs

[ Parent | on HN ]

Re:SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second ... by hhh on Sunday May 05, 2024 @03:26AM (#40262974)

If they don’t care about safety (or perceived safety), why do they spend so much time lobotomizing models for safety reasons?

[ Parent | on HN ]

Re:SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second ... by Terretta on Sunday May 05, 2024 @07:21AM (#40263997)

market reach e.g. ability to have chat app on iOS (the API is less limited)

public relations, limit the edge case nonsense 'journalists' hype so corporate execs aren't terrorized into avoiding buying

doesn't have to be as smart as it could be, it just has to be smarter than other models, so might as well file down some sharp edges for sake of above

[ Parent | on HN ]

Re:SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second ... by realusername on Sunday May 05, 2024 @10:26AM (#40265139)

Because of PR reasons. They want to avoid government legislations and pretending that they care helps

[ Parent | on HN ]

Re:SEQUOIA: Exact Llama2-70B on an RTX4090 with half-second ... by whimsicalism on Sunday May 05, 2024 @10:55AM (#40265340)

I didn’t say they don’t care about safety, merely that of the big labs they care the least or close to the least

[ Parent | on HN ]