vllm.models.muse_spark.nvidia.specialized_decode.cutedsl.oproj

Modules:

• all_reduce_dual_norm –

  The attention boundary at TP4 and TP8: the o_proj all-reduce fused with

• batt –

  The o_proj GEMM and the attention boundary (o_proj all-reduce,

• draft_all_reduce_boundary –

  The DSpark draft's residual boundary with its branch all-reduce, in one