-
|
Local execution removes dependency on connectivity, increases reliability and reduces latency. |
Beta Was this translation helpful? Give feedback.
Answered by
adamdbrw
Aug 29, 2024
Replies: 1 comment
-
|
Currently, state-of-the-art language models are too big and demanding to fit well with most of robot compute platforms. However, the technology is progressing fast and one of chief long-term goals for RAI is on-board execution. Our goal is to abstract the model vendor and hosting, thus making it easy to switch to local execution once technology allows it. |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
adamdbrw
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Currently, state-of-the-art language models are too big and demanding to fit well with most of robot compute platforms. However, the technology is progressing fast and one of chief long-term goals for RAI is on-board execution. Our goal is to abstract the model vendor and hosting, thus making it easy to switch to local execution once technology allows it.