primitive-ai

/

Qwen3.8-Flash-Next-PLE-quant

Like 4

Follow

primitive 39

Thanks for thie very good solution for a single rtx pro 6000!!
But unfortunately the hicache support for qwen3.8 architecture on vllm is not as good as sglang.
If there is a version of SSD PLE quant for sglang would be so nice !