SQ
@TpBlair just mailed you
hi @dengbuqi @larson-carter, I am also interested in unified idle GPUs into a public network and am just transferred from Petals to exo. May I ask if exo supports GPU...
> 我2块A800 80G gpu ，72核cpu，可以把cpu用满，但是显卡利用利只有10%， 都是用cpu在计算了，怎么把显卡性能用起来，显卡还有90%性能没用上，只用了10G显存，怎调？各位有办法吗？ > > 用这个参数会快一些: --optimize_rule_path ./ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu.yaml > > prompt eval count: 60 token(s) prompt eval duration: 3.1178791522979736s prompt eval rate: 19.243850408948063 tokens/s eval count: 3673...
I also met this issue. I started a private swarm with two GPU servers and built this web chat on another CPU node.
@lipere123 Hi, may I ask if you have figured this out?