vllm.models.common.ops

Ops shared across model implementations.

Modules:

• fused_allreduce_rms_norm –

  Fused all-reduce + residual-add + RMSNorm for eager model paths.

• gemm_rs_ar –

  SM100 BF16 GEMM with fused tensor-parallel reduce-scatter/all-reduce.

• sequence_parallel –