Index _ | A | B | C | D | E | F | G | H | I | K | L | M | N | P | Q | R | S | T | U | V | W | X _ __init__() (flashinfer.attention.BatchAttention method) (flashinfer.attention.BatchAttentionWithAttentionSinkWrapper method) (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.comm.AllReduceFusionOp method) (flashinfer.comm.AllReduceFusionPattern method) (flashinfer.comm.AllReduceFusionWorkspace method) (flashinfer.comm.AllReduceStrategyConfig method) (flashinfer.comm.AllReduceStrategyType method) (flashinfer.comm.CudaRTLibrary method) (flashinfer.comm.Mapping method) (flashinfer.comm.mixed_comm.MixedCommHandler method) (flashinfer.comm.mixed_comm.MixedCommMode method) (flashinfer.comm.mixed_comm.MixedCommOp method) (flashinfer.comm.mnnvl.McastGPUBuffer method) (flashinfer.comm.mnnvl.MnnvlMemory method) (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) (flashinfer.comm.MoeAlltoAll method) (flashinfer.comm.QuantizationSFLayout method) (flashinfer.comm.TRTLLMAllReduceFusionWorkspace method) (flashinfer.cute_dsl.AddRMSNormFP4QuantKernel method) (flashinfer.cute_dsl.attention.wrappers.batch_mla.BatchMLADecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper method) (flashinfer.cute_dsl.RMSNormFP4QuantKernel method) (flashinfer.decode.BatchDecodeMlaWithPagedKVCacheWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.decode.CUDAGraphBatchDecodeWithPagedKVCacheWrapper method) (flashinfer.fused_moe.B12xMoEWrapper method) (flashinfer.fused_moe.CuteDslMoEWrapper method) (flashinfer.fused_moe.RoutingMethodType method) (flashinfer.fused_moe.WeightLayout method) (flashinfer.gemm.SegmentGEMMWrapper method) (flashinfer.logits_processor.LogitsPipe method) (flashinfer.logits_processor.LogitsProcessor method) (flashinfer.logits_processor.MinP method) (flashinfer.logits_processor.Sample method) (flashinfer.logits_processor.Softmax method) (flashinfer.logits_processor.TaggedTensor method) (flashinfer.logits_processor.Temperature method) (flashinfer.logits_processor.TensorType method) (flashinfer.logits_processor.TopK method) (flashinfer.logits_processor.TopP method) (flashinfer.mla.BatchMLAPagedAttentionWrapper method) (flashinfer.pod.BatchPODWithPagedKVCacheWrapper method) (flashinfer.pod.PODWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithRaggedKVCacheWrapper method) (flashinfer.quantization.SfLayout method) (flashinfer.sparse.BlockSparseAttentionWrapper method) (flashinfer.sparse.VariableBlockSparseAttentionWrapper method) A add_rmsnorm_fp4quant() (in module flashinfer.cute_dsl) AddRMSNormFP4QuantKernel (class in flashinfer.cute_dsl) allreduce_fusion() (in module flashinfer.comm) AllReduceFusionOp (class in flashinfer.comm) AllReduceFusionPattern (class in flashinfer.comm) AllReduceFusionWorkspace (class in flashinfer.comm) AllReduceStrategyConfig (class in flashinfer.comm) AllReduceStrategyType (class in flashinfer.comm) append_paged_kv_cache() (in module flashinfer.page) append_paged_mla_kv_cache() (in module flashinfer.page) apply_llama31_rope() (in module flashinfer.rope) apply_llama31_rope_inplace() (in module flashinfer.rope) apply_llama31_rope_pos_ids() (in module flashinfer.rope) apply_llama31_rope_pos_ids_inplace() (in module flashinfer.rope) apply_rope() (in module flashinfer.rope) apply_rope_inplace() (in module flashinfer.rope) apply_rope_pos_ids() (in module flashinfer.rope) apply_rope_pos_ids_inplace() (in module flashinfer.rope) apply_rope_with_cos_sin_cache() (in module flashinfer.rope) apply_rope_with_cos_sin_cache_inplace() (in module flashinfer.rope) attention_flops() (in module flashinfer.testing) attention_flops_with_actual_seq_lens() (in module flashinfer.testing) attention_tb_per_sec() (in module flashinfer.testing) attention_tb_per_sec_with_actual_seq_lens() (in module flashinfer.testing) attention_tflops_per_sec() (in module flashinfer.testing) attention_tflops_per_sec_with_actual_seq_lens() (in module flashinfer.testing) B b12x_fused_moe() (in module flashinfer.fused_moe) B12xMoEWrapper (class in flashinfer.fused_moe) batch_deepgemm_fp8_nt_groupwise() (in module flashinfer.gemm) BatchAttention (class in flashinfer.attention) BatchAttentionWithAttentionSinkWrapper (class in flashinfer.attention) BatchDecodeMlaWithPagedKVCacheWrapper (class in flashinfer.decode) BatchDecodeWithPagedKVCacheWrapper (class in flashinfer.decode) BatchDecodeWithSharedPrefixPagedKVCacheWrapper (class in flashinfer.cascade) BatchMLADecodeCuteDSLWrapper (class in flashinfer.cute_dsl.attention.wrappers.batch_mla) BatchMLAPagedAttentionWrapper (class in flashinfer.mla) BatchPODWithPagedKVCacheWrapper (class in flashinfer.pod) BatchPrefillCuteDSLWrapper (class in flashinfer.cute_dsl.attention.wrappers.batch_prefill) BatchPrefillWithPagedKVCacheWrapper (class in flashinfer.prefill) BatchPrefillWithRaggedKVCacheWrapper (class in flashinfer.prefill) BatchPrefillWithSharedPrefixPagedKVCacheWrapper (class in flashinfer.cascade) begin_forward() (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) bench_gpu_time() (in module flashinfer.testing) bench_gpu_time_with_cuda_event() (in module flashinfer.testing) bench_gpu_time_with_cudagraph() (in module flashinfer.testing) bench_gpu_time_with_cupti() (in module flashinfer.testing) bgmv_moe() (in module flashinfer.fused_moe) bgmv_moe_expand() (in module flashinfer.fused_moe) bgmv_moe_shrink() (in module flashinfer.fused_moe) block_scale_interleave() (in module flashinfer.quantization) BlockSparseAttentionWrapper (class in flashinfer.sparse) bmm_bf16() (in module flashinfer.gemm) bmm_fp8() (in module flashinfer.gemm) bmm_mxfp8() (in module flashinfer.gemm) C can_implement_filtered_topk() (in module flashinfer.topk) chain_speculative_sampling() (in module flashinfer.sampling) checkpointing_ssu() (in module flashinfer.mamba) chunk_gated_delta_rule() (in module flashinfer.gdn_prefill) combine() (flashinfer.comm.MoeAlltoAll method) compute_fp4_swizzled_layout_sf_size() (in module flashinfer.comm) convert_to_block_layout() (in module flashinfer.fused_moe) create_allreduce_fusion_workspace() (in module flashinfer.comm) create_shared_buffer() (in module flashinfer.comm) CUDAGraphBatchDecodeWithPagedKVCacheWrapper (class in flashinfer.decode) CudaRTLibrary (class in flashinfer.comm) cudnn_batch_decode_with_kv_cache() (in module flashinfer.cudnn) (in module flashinfer.decode) cudnn_batch_prefill_with_kv_cache() (in module flashinfer.cudnn) (in module flashinfer.prefill) cute_dsl_fused_moe_nvfp4() (in module flashinfer.fused_moe) CuteDslMoEWrapper (class in flashinfer.fused_moe) cutlass_fused_moe() (in module flashinfer.fused_moe) D decode_cp_a2a_allocate_mnnvl_workspace() (in module flashinfer.comm) decode_cp_a2a_alltoall() (in module flashinfer.comm) decode_cp_a2a_init_workspace() (in module flashinfer.comm) decode_cp_a2a_workspace_size() (in module flashinfer.comm) dispatch() (flashinfer.comm.MoeAlltoAll method) E e2m1_and_ufp8sf_scale_to_float() (in module flashinfer.quantization) end_forward() (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) F fmha_v2_prefill_deepseek() (in module flashinfer.prefill) forward() (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) fp4_quantize() (in module flashinfer.quantization) fp8_blockscale_gemm_sm90() (in module flashinfer.gemm) free_shared_buffer() (in module flashinfer.comm) fused_add_rmsnorm() (in module flashinfer.norm) fused_add_rmsnorm_quant() (in module flashinfer.norm) fused_dit_gate_residual_layernorm_gamma_beta() (in module flashinfer.norm) fused_dit_gate_residual_layernorm_scale_shift() (in module flashinfer.norm) fused_dit_residual_layernorm_scale_shift() (in module flashinfer.norm) fused_qk_rmsnorm_rope() (in module flashinfer.norm) fused_rmsnorm_silu() (in module flashinfer.norm) fused_topk_deepseek() (in module flashinfer.fused_moe) G gated_delta_rule_decode() (in module flashinfer.gdn_decode) gated_delta_rule_decode_pretranspose() (in module flashinfer.gdn_decode) gated_delta_rule_mtp() (in module flashinfer.gdn_decode) gelu_and_mul() (in module flashinfer.activation) gelu_tanh_and_mul() (in module flashinfer.activation) gemm_fp8_nt_blockscaled() (in module flashinfer.gemm) gemm_fp8_nt_groupwise() (in module flashinfer.gemm) gemma_fused_add_rmsnorm() (in module flashinfer.norm) gemma_rmsnorm() (in module flashinfer.norm) get_batch_indices_positions() (in module flashinfer.page) get_combine_payload_tensor_in_workspace() (flashinfer.comm.MoeAlltoAll method) get_moe_workspace_size_per_rank() (flashinfer.comm.MoeAlltoAll static method) get_valid_tactics() (flashinfer.fused_moe.CuteDslMoEWrapper method) group_deepgemm_fp8_nt_groupwise() (in module flashinfer.gemm) group_gemm_fp8_nt_groupwise() (in module flashinfer.gemm) group_gemm_mxfp8_mxfp4_nt_groupwise() (in module flashinfer.gemm) group_gemm_nvfp4_nt_groupwise() (in module flashinfer.gemm) grouped_gemm_nt_masked() (in module flashinfer.gemm) grouped_mm_bf16() (in module flashinfer.grouped_mm) grouped_mm_fp4() (in module flashinfer.grouped_mm) grouped_mm_fp8() (in module flashinfer.grouped_mm) grouped_mm_mxfp8() (in module flashinfer.grouped_mm) H hidden_size (flashinfer.fused_moe.CuteDslMoEWrapper attribute) I interleave_moe_scales_for_sm90_mixed_gemm() (in module flashinfer.fused_moe) interleave_moe_weights_for_sm90_mixed_gemm() (in module flashinfer.fused_moe) intermediate_size (flashinfer.fused_moe.CuteDslMoEWrapper attribute) is_cute_dsl_available() (in module flashinfer.cute_dsl) K kernel() (flashinfer.cute_dsl.AddRMSNormFP4QuantKernel method) (flashinfer.cute_dsl.RMSNormFP4QuantKernel method) L layernorm() (in module flashinfer.norm) LogitsPipe (class in flashinfer.logits_processor) LogitsProcessor (class in flashinfer.logits_processor) M Mapping (class in flashinfer.comm) max_num_tokens (flashinfer.fused_moe.CuteDslMoEWrapper attribute) McastGPUBuffer (class in flashinfer.comm.mnnvl) merge_state() (in module flashinfer.cascade) merge_state_in_place() (in module flashinfer.cascade) merge_states() (in module flashinfer.cascade) min_p_sampling_from_probs() (in module flashinfer.sampling) MinP (class in flashinfer.logits_processor) MixedCommHandler (class in flashinfer.comm.mixed_comm) MixedCommMode (class in flashinfer.comm.mixed_comm) MixedCommOp (class in flashinfer.comm.mixed_comm) mla_rope_quantize_fp8() (in module flashinfer.rope) mm_bf16() (in module flashinfer.gemm) mm_fp4() (in module flashinfer.gemm) mm_fp8() (in module flashinfer.gemm) mm_M1_16_K6144_N256() (in module flashinfer.gemm) mm_M1_16_K7168_N128() (in module flashinfer.gemm) mm_M1_16_K7168_N256() (in module flashinfer.gemm) mm_mxfp8() (in module flashinfer.gemm) MNNVLAllReduceFusionWorkspace (class in flashinfer.comm) MnnvlMemory (class in flashinfer.comm.mnnvl) moe_a2a_combine() (in module flashinfer.comm) moe_a2a_dispatch() (in module flashinfer.comm) moe_a2a_get_workspace_size_per_rank() (in module flashinfer.comm) moe_a2a_initialize() (in module flashinfer.comm) moe_a2a_sanitize_expert_ids() (in module flashinfer.comm) moe_a2a_wrap_payload_tensor_in_workspace() (in module flashinfer.comm) moe_gemm_mxfp8_nt_groupwise() (in module flashinfer.grouped_mm) MoeAlltoAll (class in flashinfer.comm) mpi_barrier() (in module flashinfer.comm.trtllm_mnnvl_ar) MultiLevelCascadeAttentionWrapper (class in flashinfer.cascade) mxfp4_dequantize() (in module flashinfer.quantization) mxfp4_dequantize_host() (in module flashinfer.quantization) mxfp4_quantize() (in module flashinfer.quantization) mxfp4_quantize_cute_dsl() (in module flashinfer.quantization.kernels.mxfp4_quantize) mxfp8_dequantize_host() (in module flashinfer.quantization) mxfp8_quantize() (in module flashinfer.quantization) mxfp8_quantize_cute_dsl() (in module flashinfer.quantization.kernels.mxfp8_quantize) N num_experts (flashinfer.fused_moe.CuteDslMoEWrapper attribute) nvfp4_attention_sm120_fwd() (in module flashinfer.nvfp4_attention_sm120) nvfp4_attention_sm120_quantize_qkv() (in module flashinfer.nvfp4_attention_sm120) nvfp4_batched_quantize() (in module flashinfer.quantization) nvfp4_kv_dequantize() (in module flashinfer.quantization) nvfp4_kv_quantize() (in module flashinfer.quantization) nvfp4_quantize() (in module flashinfer.quantization) nvfp4_quantize_cute_dsl() (in module flashinfer.quantization.kernels.nvfp4_quantize) nvfp4_quantize_paged_kv_cache() (in module flashinfer.quantization) nvfp4_quantize_per_token_cute_dsl() (in module flashinfer.quantization.kernels.nvfp4_quantize) P pack_strided_memory() (in module flashinfer.comm) packbits() (in module flashinfer.quantization) plan() (flashinfer.attention.BatchAttention method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_mla.BatchMLADecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper method) (flashinfer.decode.BatchDecodeMlaWithPagedKVCacheWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.mla.BatchMLAPagedAttentionWrapper method) (flashinfer.pod.BatchPODWithPagedKVCacheWrapper method) (flashinfer.pod.PODWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithRaggedKVCacheWrapper method) (flashinfer.sparse.BlockSparseAttentionWrapper method) (flashinfer.sparse.VariableBlockSparseAttentionWrapper method) PODWithPagedKVCacheWrapper (class in flashinfer.pod) Q QuantizationSFLayout (class in flashinfer.comm) R reorder_rows_for_gated_act_gemm() (in module flashinfer.fused_moe) reset_workspace_buffer() (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.decode.BatchDecodeMlaWithPagedKVCacheWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.gemm.SegmentGEMMWrapper method) (flashinfer.pod.PODWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithRaggedKVCacheWrapper method) (flashinfer.sparse.BlockSparseAttentionWrapper method) (flashinfer.sparse.VariableBlockSparseAttentionWrapper method) rmsnorm() (in module flashinfer.norm) rmsnorm_fp4quant() (in module flashinfer.cute_dsl) rmsnorm_quant() (in module flashinfer.norm) RMSNormFP4QuantKernel (class in flashinfer.cute_dsl) rope_quantize_fp8() (in module flashinfer.rope) rope_quantize_fp8_append_paged_kv_cache() (in module flashinfer.rope) RoutingMethodType (class in flashinfer.fused_moe) run() (flashinfer.attention.BatchAttention method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_mla.BatchMLADecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper method) (flashinfer.decode.BatchDecodeMlaWithPagedKVCacheWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.fused_moe.B12xMoEWrapper method) (flashinfer.fused_moe.CuteDslMoEWrapper method) (flashinfer.gemm.SegmentGEMMWrapper method) (flashinfer.mla.BatchMLAPagedAttentionWrapper method) (flashinfer.pod.BatchPODWithPagedKVCacheWrapper method) (flashinfer.pod.PODWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithRaggedKVCacheWrapper method) (flashinfer.sparse.BlockSparseAttentionWrapper method) (flashinfer.sparse.VariableBlockSparseAttentionWrapper method) run_mixed_comm() (in module flashinfer.comm.mixed_comm) S Sample (class in flashinfer.logits_processor) sampling_from_logits() (in module flashinfer.sampling) sampling_from_probs() (in module flashinfer.sampling) scaled_fp4_grouped_quantize() (in module flashinfer.quantization) segment_packbits() (in module flashinfer.quantization) SegmentGEMMWrapper (class in flashinfer.gemm) selective_state_update() (in module flashinfer.mamba) set_seed() (in module flashinfer.testing) SfLayout (class in flashinfer.quantization) shuffle_matrix_a() (in module flashinfer.quantization) shuffle_matrix_sf_a() (in module flashinfer.quantization) silu_and_mul() (in module flashinfer.activation) silu_and_mul_scaled_nvfp4_experts_quantize() (in module flashinfer.activation) single_decode_with_kv_cache() (in module flashinfer.decode) single_decode_with_kv_cache_with_jit_module() (in module flashinfer.decode) single_prefill_with_kv_cache() (in module flashinfer.prefill) single_prefill_with_kv_cache_return_lse() (in module flashinfer.prefill) single_prefill_with_kv_cache_with_jit_module() (in module flashinfer.prefill) sleep_after_kernel_run() (in module flashinfer.testing) Softmax (class in flashinfer.logits_processor) softmax() (in module flashinfer.sampling) split_device_green_ctx() (in module flashinfer.green_ctx) split_device_green_ctx_by_sm_count() (in module flashinfer.green_ctx) T TaggedTensor (class in flashinfer.logits_processor) Temperature (class in flashinfer.logits_processor) TensorType (class in flashinfer.logits_processor) tgv_gemm_sm100() (in module flashinfer.gemm) tinygemm_bf16() (in module flashinfer.gemm) top_k (flashinfer.fused_moe.CuteDslMoEWrapper attribute) top_k() (in module flashinfer) top_k_mask_logits() (in module flashinfer.sampling) top_k_page_table_transform() (in module flashinfer) top_k_ragged_transform() (in module flashinfer) top_k_renorm_probs() (in module flashinfer.sampling) top_k_sampling_from_probs() (in module flashinfer.sampling) top_k_top_p_sampling_from_logits() (in module flashinfer.sampling) top_k_top_p_sampling_from_probs() (in module flashinfer.sampling) top_p_renorm_probs() (in module flashinfer.sampling) top_p_sampling_from_probs() (in module flashinfer.sampling) TopK (class in flashinfer.logits_processor) TopP (class in flashinfer.logits_processor) trtllm_allreduce_fusion() (in module flashinfer.comm) trtllm_batch_context_with_kv_cache() (in module flashinfer.prefill) trtllm_batch_decode_with_kv_cache() (in module flashinfer.decode) trtllm_batch_decode_with_kv_cache_mla() (in module flashinfer.mla) trtllm_bf16_moe() (in module flashinfer.fused_moe) trtllm_bf16_routed_moe() (in module flashinfer.fused_moe) trtllm_create_ipc_workspace_for_all_reduce() (in module flashinfer.comm) trtllm_create_ipc_workspace_for_all_reduce_fusion() (in module flashinfer.comm) trtllm_custom_all_reduce() (in module flashinfer.comm) trtllm_destroy_ipc_workspace_for_all_reduce() (in module flashinfer.comm) trtllm_destroy_ipc_workspace_for_all_reduce_fusion() (in module flashinfer.comm) trtllm_fmha_v2_prefill() (in module flashinfer.prefill) trtllm_fp4_block_scale_moe() (in module flashinfer.fused_moe) trtllm_fp4_block_scale_routed_moe() (in module flashinfer.fused_moe) trtllm_fp8_block_scale_moe() (in module flashinfer.fused_moe) trtllm_fp8_block_scale_routed_moe() (in module flashinfer.fused_moe) trtllm_fp8_per_tensor_scale_moe() (in module flashinfer.fused_moe) trtllm_lamport_initialize() (in module flashinfer.comm) trtllm_lamport_initialize_all() (in module flashinfer.comm) trtllm_mnnvl_all_reduce() (in module flashinfer.comm.trtllm_mnnvl_ar) trtllm_mnnvl_allreduce() (in module flashinfer.comm.trtllm_mnnvl_ar) trtllm_mnnvl_fused_allreduce_add_rmsnorm() (in module flashinfer.comm.trtllm_mnnvl_ar) trtllm_mnnvl_fused_allreduce_add_rmsnorm_quant() (in module flashinfer.comm.trtllm_mnnvl_ar) trtllm_mnnvl_fused_allreduce_rmsnorm() (in module flashinfer.comm.trtllm_mnnvl_ar) trtllm_moe_allreduce_fusion() (in module flashinfer.comm) trtllm_moe_finalize_allreduce_fusion() (in module flashinfer.comm) trtllm_mxint4_block_scale_moe() (in module flashinfer.fused_moe) trtllm_mxint4_block_scale_routed_moe() (in module flashinfer.fused_moe) trtllm_ragged_attention_deepseek() (in module flashinfer.prefill) TRTLLMAllReduceFusionWorkspace (class in flashinfer.comm) U use_cuda_graph (flashinfer.fused_moe.CuteDslMoEWrapper attribute) V VariableBlockSparseAttentionWrapper (class in flashinfer.sparse) vllm_all_reduce() (in module flashinfer.comm) vllm_dispose() (in module flashinfer.comm) vllm_get_graph_buffer_ipc_meta() (in module flashinfer.comm) vllm_init_custom_ar() (in module flashinfer.comm) vllm_meta_size() (in module flashinfer.comm) vllm_register_buffer() (in module flashinfer.comm) vllm_register_graph_buffers() (in module flashinfer.comm) W WeightLayout (class in flashinfer.fused_moe) X xqa() (in module flashinfer.xqa) xqa_batch_decode_with_kv_cache() (in module flashinfer.decode) xqa_batch_decode_with_kv_cache_mla() (in module flashinfer.mla) xqa_mla() (in module flashinfer.xqa)