The context window is 16 characters. Talking about tokens per second is meaningless.
show comments
cadamsdotcom
Transformers scale poorly vs. context window size and parameter count.
Which means really impressive when those N’s are small!
I’m but a pundit in this area so don’t know much. But one wonders if there’s a future in burning larger models to FPGAs - whether big enough FPGAs exist (or can be built), and whether locating specialized compute right with the memory it needs can speed things up.
Likely would need a lot of algorithm parallelism work that’d translate back to CPUs/GPUs.
The context window is 16 characters. Talking about tokens per second is meaningless.
Transformers scale poorly vs. context window size and parameter count.
Which means really impressive when those N’s are small!
I’m but a pundit in this area so don’t know much. But one wonders if there’s a future in burning larger models to FPGAs - whether big enough FPGAs exist (or can be built), and whether locating specialized compute right with the memory it needs can speed things up.
Likely would need a lot of algorithm parallelism work that’d translate back to CPUs/GPUs.
See also:
https://rits.shanghai.nyu.edu/ai/karpathys-microgpt-on-fpga-...
TL;DR: The CPU implementation was 71x faster than the FPGA.
Note: model has only 4192 parameters.