vllm.v1.attention.backends.mla ¶

Modules:

• flashinfer_mla_sparse –

  FlashInfer MLA Sparse Attention Backend.

• flashmla_sparse –
• indexer –
• prefill –
• rocm_aiter_mla –
• rocm_aiter_mla_sparse –
• sparse_swa –
• sparse_utils –

  Utility functions for sparse MLA backends.

• tokenspeed_mla –

  TokenSpeed CuTe DSL MLA decode backend (Blackwell, FP8 KV cache only).