implementation detail · filed under software implementation
vLLM tensor parallelism
A vLLM serving configuration that distributes inference across eight GPUs using tensor parallelism.
Also called vLLM at --tensor-parallel-size 8.
- source
- 1
- models
- 2
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
Xiaomi’s model card serves it with vLLM at --tensor-parallel-size 8
usedunclearin MiMo-V2.6-ProXiaomi
Filed alongside
Other methods under software implementation :: inference engine.