revision 2: correct the 310P artifact, add the compile-source ONNX, align deployments with org convention

add ascend_310b deployment (README.md)

add ascend_310b deployment (inference_manifest.json)

add ascend_310b deployment (silero_vad_v6_310b_fp16.om)

Update Silero VAD Ascend 310P OM

Initial upload: dual deployment (ascend_310p OM + torch_cpu v6 ONNX) with stream state contract

Initial model card: standalone Silero VAD bundle (ascend_310p + torch_cpu)

initial commit