vllm.v1.attention.ops ¶

Modules:

• chunked_prefill_paged_decode –
• common –
• dcp_alltoall –

  DCP All-to-All communication backend for attention.

• flashmla –
• merge_attn_states –
• rocm_aiter_mla_sparse –
• triton_attention_helpers –

  Shared @triton.jit helpers used by the unified attention kernel

• triton_decode_attention –

  Memory-efficient attention for decoding.

• triton_prefill_attention –

  Memory-efficient attention for prefill.

• triton_reshape_and_cache_flash –
• triton_turboquant_decode –

  Triton fused TurboQuant decode attention.

• triton_turboquant_store –

  Fused Triton kernels for TurboQuant KV store.

• triton_unified_attention –
• vit_attn_wrappers –

  This file contains ops for ViT attention to be compatible with torch.compile

• xpu_mla_sparse –