A vLLM (0.12.0) out-of-tree platform plugin that enables running vLLM on NPU (Ascend/torch_npu).
最近更新: 28天前Omni_Infer is a suite of inference accelerators designed for the Ascend NPU platform, offering native support and an expanding feature set.
最近更新: 6个月前cann-ops-adv,是基于昇腾硬件的融合算子库(adv表示advanced)。
最近更新: 7个月前