Kyle
@darsnack.bsky.social
NeuroAI Scholar @ CSHL
https://www.darsnack.info
Previously maintaining FluxML to procrastinate Previously EE PhD at UW-Madison, comp. eng. / math at Rose-Hulman
https://www.darsnack.info
Previously maintaining FluxML to procrastinate Previously EE PhD at UW-Madison, comp. eng. / math at Rose-Hulman
I guess the take-away is that frozen LLMs can be heavily quantized down to 2-bits which makes any classic approach (CPU, GPU, ASIC) more competitive in terms of power. 1/
November 12, 2024 at 6:25 PM
I guess the take-away is that frozen LLMs can be heavily quantized down to 2-bits which makes any classic approach (CPU, GPU, ASIC) more competitive in terms of power. 1/
Will New York ever be the same
October 15, 2024 at 10:29 PM
Will New York ever be the same