implementation detail · not yet filed
Limit maximum concurrent sequences to 256
Also called keep --max-num-seqs 256.
- source
- 1
- model
- 1
- 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.
Mamba-cache capacity error at startup: keep --max-num-seqs 256.
usedinference servingin Qwen3.8-Flash-NextvLLM recipe authors