vllm.v1.attention.backends ¶

Modules:

• cpu_attn –
• fa_utils –
• flash_attn –

  Attention layer with FlashAttention.

• flash_attn_diffkv –

  Attention layer with FlashAttention.

• flashinfer –

  Attention layer with FlashInfer.

• flex_attention –

  Attention layer with FlexAttention.

• gdn_attn –

  Backend for GatedDeltaNet attention.

• mamba2_attn –
• mamba_attn –
• mla –
• registry –

  Attention backend registry

• rocm_aiter_fa –

  Attention layer with AiterFlashAttention.

• rocm_aiter_unified_attn –

  Attention layer with PagedAttention and Triton prefix prefill.

• rocm_attn –

  Attention layer with PagedAttention and Triton prefix prefill.

• triton_attn –

  High-Performance Triton-only Attention layer.

• turboquant_attn –

  TurboQuant attention backend for vLLM.

• utils –