A vLLM (0.12.0) out-of-tree platform plugin that enables running vLLM on NPU (Ascend/torch_npu).
最近更新: 5小时前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: 2个月前OmniOperator operator acceleration is implemented using native code (C/C++) to optimize big data SQL operators.
最近更新: 5个月前
