Index _ | A | B | C | D | E | F | G | H | I | K | L | M | N | P | Q | R | S | T | U | V | W | X _ __call__() (flashinfer.fused_moe.MoELayer method) __init__() (flashinfer.attention.BatchAttention method) (flashinfer.attention.BatchAttentionWithAttentionSinkWrapper method) (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (flashinfer.cake_rmsnorm_train.CakeRMSNormFunction method) (flashinfer.CakeFmhaRequestOrderedDecodePlan 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.mnnvl_cutedsl_ag.MNNVLCuteDSLAllGatherDualNormWorkspace method) (flashinfer.comm.mnnvl_cutedsl_ag.MNNVLCuteDSLAllGatherNormQuantWorkspace method) (flashinfer.comm.mnnvl_cutedsl_ar.MNNVLCuteDSLAllReduceFusionWorkspace method) (flashinfer.comm.mnnvl_cutedsl_moe_finalize.MNNVLCuteDSLMoEFinalizeAllReduceWorkspace method) (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) (flashinfer.comm.MoeAlltoAll method) (flashinfer.comm.PcieIpcAllGatherLaunchConfig method) (flashinfer.comm.PcieIpcAllGatherVariant method) (flashinfer.comm.PcieIpcAllGatherWorkspace method) (flashinfer.comm.PcieIpcAllReduceWorkspace method) (flashinfer.comm.PcieIpcReduceScatterLaunchConfig method) (flashinfer.comm.PcieIpcReduceScatterVariant method) (flashinfer.comm.PcieIpcReduceScatterWorkspace method) (flashinfer.comm.QuantizationSFLayout method) (flashinfer.comm.TRTLLMAllReduceFusionWorkspace method) (flashinfer.comm.UlyssesBackendDecision method) (flashinfer.comm.UlyssesCommunicator method) (flashinfer.comm.UlyssesRankTopology method) (flashinfer.comm.UlyssesWorkspace method) (flashinfer.cute_dsl.AddRMSNormFP4QuantKernel method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodePagedCuteDSLWrapper 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.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.decode.CUDAGraphBatchDecodeWithPagedKVCacheWrapper method) (flashinfer.fused_moe.AlphaMoERoutePlan method) (flashinfer.fused_moe.B12xMoEWrapper method) (flashinfer.fused_moe.BackendOptions method) (flashinfer.fused_moe.BGMVMoECakePlan method) (flashinfer.fused_moe.BGMVMoEPortablePlan method) (flashinfer.fused_moe.CakeWarpDecodeConfig method) (flashinfer.fused_moe.CakeWarpDecodeRunner method) (flashinfer.fused_moe.CudnnFrostBf16Config method) (flashinfer.fused_moe.CudnnFrostMxfp8Config method) (flashinfer.fused_moe.CudnnFrostMxfp8Mxfp4Config method) (flashinfer.fused_moe.CudnnFrostNvfp4Config method) (flashinfer.fused_moe.CuteDslBf16MoEWrapper method) (flashinfer.fused_moe.CuteDslMoEWrapper method) (flashinfer.fused_moe.CuteDslMxfp8Mxfp4MoEWrapper method) (flashinfer.fused_moe.CuTileBf16Config method) (flashinfer.fused_moe.CuTileBf16Runner method) (flashinfer.fused_moe.CuTileFp8PerTensorBf16Config method) (flashinfer.fused_moe.CuTileFp8PerTensorBf16Runner method) (flashinfer.fused_moe.CuTileFp8PerTensorConfig method) (flashinfer.fused_moe.CuTileFp8PerTensorRunner method) (flashinfer.fused_moe.CuTileMxfp4Bf16Config method) (flashinfer.fused_moe.CuTileMxfp4Bf16Runner method) (flashinfer.fused_moe.CuTileMxfp4Config method) (flashinfer.fused_moe.CuTileMxfp4Mxfp8Config method) (flashinfer.fused_moe.CuTileMxfp4Mxfp8Runner method) (flashinfer.fused_moe.CuTileMxfp4Runner method) (flashinfer.fused_moe.CuTileMxfp8Bf16Config method) (flashinfer.fused_moe.CuTileMxfp8Bf16Runner method) (flashinfer.fused_moe.CuTileMxfp8Config method) (flashinfer.fused_moe.CuTileMxfp8Runner method) (flashinfer.fused_moe.CuTileNvfp4Bf16Config method) (flashinfer.fused_moe.CuTileNvfp4Bf16Runner method) (flashinfer.fused_moe.CuTileNvfp4Config method) (flashinfer.fused_moe.CuTileNvfp4Runner method) (flashinfer.fused_moe.ExecutionConfig method) (flashinfer.fused_moe.ExpertConfig method) (flashinfer.fused_moe.MoEActivationPack method) (flashinfer.fused_moe.MoEConfig method) (flashinfer.fused_moe.MoEFinalizeConfig method) (flashinfer.fused_moe.MoELayer method) (flashinfer.fused_moe.MoEWeightPack method) (flashinfer.fused_moe.PrimsTsConfig method) (flashinfer.fused_moe.PrimsTsRunner method) (flashinfer.fused_moe.QuantConfig method) (flashinfer.fused_moe.QuantFormat method) (flashinfer.fused_moe.RoutingConfig method) (flashinfer.fused_moe.RoutingMethodType method) (flashinfer.fused_moe.TrtllmGenRoutingResult method) (flashinfer.fused_moe.WeightLayout method) (flashinfer.gemm.SegmentGEMMWrapper method) (flashinfer.kda_prefill.RecurrentKDAPrefillWorkspace 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.mla.DSV4HCAMetadata method) (flashinfer.mla.SparseMLASm120DecodeConfig method) (flashinfer.msa_ops.MSASparseAttentionWorkspace method) (flashinfer.pod.BatchPODWithPagedKVCacheWrapper method) (flashinfer.pod.PODWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithRaggedKVCacheWrapper method) (flashinfer.quantization.NVFP44Over6Config method) (flashinfer.quantization.NVFP44Over6ErrMode 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) all_gather_gated_residual_next_rmsnorm() (in module flashinfer.comm.mnnvl_cutedsl_ag) all_gather_matmul() (in module flashinfer.comm) all_gather_rmsnorm_fp4_quantize() (in module flashinfer.comm.mnnvl_cutedsl_ag) alloc_scratchpad() (in module flashinfer.fused_moe) allocate_alphamoe_route_plan() (in module flashinfer.fused_moe) 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) alphamoe_fp8_block_scale_aligned_moe() (in module flashinfer.fused_moe) alphamoe_fused_router() (in module flashinfer.fused_moe) alphamoe_interleave_gated_weights() (in module flashinfer.fused_moe) alphamoe_nvfp4_aligned_moe() (in module flashinfer.fused_moe) alphamoe_nvfp4_routed_moe() (in module flashinfer.fused_moe) AlphaMoERoutePlan (class in flashinfer.fused_moe) 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) (in module flashinfer.fused_moe.cute_dsl) B12xMoEWrapper (class in flashinfer.fused_moe) backend (flashinfer.comm.MNNVLAllReduceFusionWorkspace property) (flashinfer.comm.TRTLLMAllReduceFusionWorkspace property) BackendOptions (class in flashinfer.fused_moe) batch_deepgemm_fp8_nt_groupwise() (in module flashinfer.gemm) BatchAttention (class in flashinfer.attention) BatchAttentionWithAttentionSinkWrapper (class in flashinfer.attention) BatchDecodeCuteDSLWrapper (class in flashinfer.cute_dsl.attention.wrappers.batch_decode) BatchDecodePagedCuteDSLWrapper (class in flashinfer.cute_dsl.attention.wrappers.batch_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) BatchPrefillWithCausalBidirectionalRangesWrapper (class in flashinfer.attention) BatchPrefillWithPagedKVCacheWrapper (class in flashinfer.prefill) BatchPrefillWithRaggedKVCacheWrapper (class in flashinfer.prefill) BatchPrefillWithSharedPrefixPagedKVCacheWrapper (class in flashinfer.cascade) begin_forward() (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (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_gemm1_lora_delta() (in module flashinfer.fused_moe) bgmv_moe_gemm2_lora_delta() (in module flashinfer.fused_moe) bgmv_moe_shrink() (in module flashinfer.fused_moe) BGMVMoECakePlan (class in flashinfer.fused_moe) BGMVMoEPortablePlan (class in 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) bsa_attn_sm100_blk128_fwd() (in module flashinfer.cute_dsl.sparse.bsa_attn_sm100_blk128) bsa_attn_sm100_blk64_fwd() (in module flashinfer.cute_dsl.sparse.bsa_attn_sm100_blk64) bsa_attn_sm120_blk64_fwd() (in module flashinfer.cute_dsl.sparse.bsa_attn_sm120) bsa_attn_sm120_blk64_sage_fwd() (in module flashinfer.cute_dsl.sparse.bsa_attn_sm120) bytes_per_token (flashinfer.mla.SparseMLASm120DecodeConfig attribute) C cake_batch_context_with_kv_cache() (in module flashinfer.cake_fmha) cake_batch_decode_with_kv_cache() (in module flashinfer.cake_fmha) cake_dsv4_nvfp4_kv_rope_quantize_insert() (in module flashinfer.mla) cake_dsv4_nvfp4_rope_quantize_insert() (in module flashinfer.mla) cake_fmha_manifest() (in module flashinfer.cake_fmha) cake_grouped_mm() (in module flashinfer.experimental.cake_moe_grouped_gemm) cake_rmsnorm() (in module flashinfer.cake_rmsnorm_train) cake_rmsnorm_train_backward() (in module flashinfer.cake_rmsnorm_train) cake_rmsnorm_train_backward_workspace() (in module flashinfer.cake_rmsnorm_train) cake_rmsnorm_train_backward_workspace_bytes() (in module flashinfer.cake_rmsnorm_train) cake_rmsnorm_train_forward() (in module flashinfer.cake_rmsnorm_train) cake_sampling_route() (in module flashinfer.cake_sampling) cake_sparse_mla_sm120_dsv4_nvfp4_decode() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_plan_head_tiles() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_plan_prefill() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_plan_splits() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_prefill() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_scratch_bytes() (in module flashinfer.mla) cake_sparse_mla_sm120_dsv4_nvfp4_select_kernel() (in module flashinfer.mla) CakeFmhaRequestOrderedDecodePlan (class in flashinfer) CakeRMSNormFunction (class in flashinfer.cake_rmsnorm_train) CakeWarpDecodeConfig (class in flashinfer.fused_moe) CakeWarpDecodeRunner (class in flashinfer.fused_moe) can_implement_filtered_topk() (in module flashinfer.topk) chain_speculative_sampling() (in module flashinfer.sampling) checkpoint_prepare() (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) (flashinfer.comm.MoeAlltoAll method), [1] (flashinfer.comm.TRTLLMAllReduceFusionWorkspace method) checkpoint_restore() (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) (flashinfer.comm.MoeAlltoAll method), [1] (flashinfer.comm.TRTLLMAllReduceFusionWorkspace method) checkpointing_ssu() (in module flashinfer.mamba) choose_stage1() (in module flashinfer.cake_sampling) choose_stage23() (in module flashinfer.cake_sampling) chunk_gated_delta_product() (in module flashinfer.gdp_prefill) chunk_gated_delta_rule() (in module flashinfer.gdn_prefill) chunk_gated_delta_rule2() (in module flashinfer.gdn2_prefill) close() (flashinfer.comm.UlyssesCommunicator method) combine() (flashinfer.comm.MoeAlltoAll method) compute_fp4_swizzled_layout_sf_size() (in module flashinfer.comm) compute_paged_mqa_logits_schedule() (in module flashinfer) concat_mla_k() (in module flashinfer.concat_ops) concat_mla_kv_quant_fp8() (in module flashinfer.mla_kv_pack) convert_compressed_page_aligned_sparse_indices_to_hca_metadata() (in module flashinfer.mla) convert_to_block_layout() (in module flashinfer.fused_moe) create_allreduce_fusion_workspace() (in module flashinfer.comm) create_shared_buffer() (in module flashinfer.comm) create_workspace() (flashinfer.comm.UlyssesCommunicator method) 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) cudnn_chunk_gated_delta_product() (in module flashinfer.cudnn) cudnn_chunk_gated_delta_rule() (in module flashinfer.cudnn) cudnn_chunk_gated_delta_rule2() (in module flashinfer.cudnn) cudnn_recurrent_kda() (in module flashinfer.cudnn) CudnnFrostBf16Config (class in flashinfer.fused_moe) CudnnFrostMxfp8Config (class in flashinfer.fused_moe) CudnnFrostMxfp8Mxfp4Config (class in flashinfer.fused_moe) CudnnFrostNvfp4Config (class in flashinfer.fused_moe) cute_dsl_fused_moe() (in module flashinfer.fused_moe) cute_dsl_fused_moe_bf16() (in module flashinfer.fused_moe) (in module flashinfer.fused_moe.cute_dsl) cute_dsl_fused_moe_mxfp8_mxfp4() (in module flashinfer.fused_moe) (in module flashinfer.fused_moe.cute_dsl) cute_dsl_fused_moe_nvfp4() (in module flashinfer.fused_moe) (in module flashinfer.fused_moe.cute_dsl) cute_dsl_hca_decode() (in module flashinfer.cute_dsl.attention.wrappers.batch_hca) CuteDslBf16MoEWrapper (class in flashinfer.fused_moe) CuteDslMoEWrapper (class in flashinfer.fused_moe) CuteDslMxfp8Mxfp4MoEWrapper (class in flashinfer.fused_moe) CuTileBf16Config (class in flashinfer.fused_moe) CuTileBf16Runner (class in flashinfer.fused_moe) CuTileFp8PerTensorBf16Config (class in flashinfer.fused_moe) CuTileFp8PerTensorBf16Runner (class in flashinfer.fused_moe) CuTileFp8PerTensorConfig (class in flashinfer.fused_moe) CuTileFp8PerTensorRunner (class in flashinfer.fused_moe) CuTileMxfp4Bf16Config (class in flashinfer.fused_moe) CuTileMxfp4Bf16Runner (class in flashinfer.fused_moe) CuTileMxfp4Config (class in flashinfer.fused_moe) CuTileMxfp4Mxfp8Config (class in flashinfer.fused_moe) CuTileMxfp4Mxfp8Runner (class in flashinfer.fused_moe) CuTileMxfp4Runner (class in flashinfer.fused_moe) CuTileMxfp8Bf16Config (class in flashinfer.fused_moe) CuTileMxfp8Bf16Runner (class in flashinfer.fused_moe) CuTileMxfp8Config (class in flashinfer.fused_moe) CuTileMxfp8Runner (class in flashinfer.fused_moe) CuTileNvfp4Bf16Config (class in flashinfer.fused_moe) CuTileNvfp4Bf16Runner (class in flashinfer.fused_moe) CuTileNvfp4Config (class in flashinfer.fused_moe) CuTileNvfp4Runner (class in flashinfer.fused_moe) cutlass_fused_moe() (in module flashinfer.fused_moe) D d_qk (flashinfer.mla.SparseMLASm120DecodeConfig attribute) decide_ulysses_backend() (in module flashinfer.comm) 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_lse_reduce() (in module flashinfer.comm) decode_cp_a2a_lse_reduce_create_workspace() (in module flashinfer.comm) decode_cp_a2a_lse_reduce_workspace_size() (in module flashinfer.comm) decode_cp_a2a_workspace_size() (in module flashinfer.comm) destroy() (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) (flashinfer.comm.TRTLLMAllReduceFusionWorkspace method) dispatch() (flashinfer.comm.MoeAlltoAll method) dispose_ulysses_a2a() (in module flashinfer.comm) dsv41_fp4_quantize_append_sparse_mla_cache() (in module flashinfer.mla) dsv41_fp4_quantize_pack_sparse_mla_cache() (in module flashinfer.mla) dsv41_fp8_quantize_append_sparse_mla_cache() (in module flashinfer.mla) dsv41_fp8_quantize_pack_sparse_mla_cache() (in module flashinfer.mla) DSV4HCAMetadata (class in flashinfer.mla) E e2m1_and_ufp8sf_scale_to_float() (in module flashinfer.quantization) empty_per_locality_domain() (in module flashinfer.green_ctx) eplb_gathered_stats (flashinfer.comm.MoeAlltoAll property) ExecutionConfig (class in flashinfer.fused_moe) ExpertConfig (class in flashinfer.fused_moe) extra_page_block_sizes (flashinfer.mla.SparseMLASm120DecodeConfig attribute) F fmha_v2_prefill_deepseek() (in module flashinfer.prefill) fmha_v2_prefill_sm120() (in module flashinfer.prefill) forward() (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) forward_return_lse() (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) fp4_linear() (in module flashinfer.gemm) fp4_linear_swiglu() (in module flashinfer.gemm) fp4_paged_mqa_logits() (in module flashinfer), [1] fp4_qkv_qknorm_rope() (in module flashinfer.gemm) fp4_quantize() (in module flashinfer.quantization) fp8_blockscale_gemm_sm90() (in module flashinfer.gemm) fp8_linear() (in module flashinfer.gemm) fp8_linear_swiglu() (in module flashinfer.gemm) fp8_paged_mqa_logits() (in module flashinfer), [1] fp8_qkv_qknorm_rope() (in module flashinfer.gemm) free_shared_buffer() (in module flashinfer.comm) fused_add_rmsnorm() (in module flashinfer.norm) fused_add_rmsnorm_fp8_block_quant() (in module flashinfer.norm) fused_add_rmsnorm_quant() (in module flashinfer.norm) fused_dit_gate_residual_layernorm_gamma_beta() (in module flashinfer.diffusion_ops) (in module flashinfer.norm) fused_dit_gate_residual_layernorm_scale_shift() (in module flashinfer.diffusion_ops) (in module flashinfer.norm) fused_dit_residual_layernorm_scale_shift() (in module flashinfer.diffusion_ops) (in module flashinfer.norm) fused_kda_decode() (in module flashinfer.kda_decode) fused_qk_rmsnorm_rope() (in module flashinfer.diffusion_ops) (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) gated_delta_rule_replayssm_commit() (in module flashinfer.gdn_decode) gated_residual_next_rmsnorm() (in module flashinfer.norm) gated_residual_rmsnorm() (in module flashinfer.norm) gather_heads() (flashinfer.comm.UlyssesCommunicator method) gather_output_head_chunk() (flashinfer.comm.UlyssesCommunicator method) gdn_fused_decode_step() (in module flashinfer) gdn_fused_decode_step_supported() (in module flashinfer) gelu_and_mul() (in module flashinfer.activation) gelu_tanh_and_mul() (in module flashinfer.activation) gemm_alpha_beta() (in module flashinfer.gemm) 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_cake_fmha_module() (in module flashinfer.cake_fmha) get_combine_payload_tensor_in_workspace() (flashinfer.comm.MoeAlltoAll method) get_dcp_spec_counter_bytes() (in module flashinfer) get_dcp_spec_workspace_size_bytes() (in module flashinfer) get_locality_domain_count() (in module flashinfer.green_ctx) get_moe_workspace_size_per_rank() (flashinfer.comm.MoeAlltoAll static method) get_pcie_ipc_all_gather_launch_config() (in module flashinfer.comm) get_pcie_ipc_launch_config() (in module flashinfer.comm) get_pcie_ipc_reduce_scatter_launch_config() (in module flashinfer.comm) get_required_buffer_size_bytes() (flashinfer.comm.MNNVLAllReduceFusionWorkspace static method) get_scratchpad_size_bytes() (in module flashinfer.fused_moe) get_valid_tactics() (flashinfer.fused_moe.CuteDslMoEWrapper method) (flashinfer.fused_moe.CuteDslMxfp8Mxfp4MoEWrapper method) get_workspace() (flashinfer.comm.MoeAlltoAll class method) group_deepgemm_fp8_nt_groupwise() (in module flashinfer.gemm) group_gemm_fp8_nt_groupwise() (in module flashinfer.gemm) group_gemm_fp8_nt_groupwise_contiguous() (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_dgrad() (in module flashinfer.experimental.cake_moe_grouped_gemm) grouped_gemm_fwd() (in module flashinfer.experimental.cake_moe_grouped_gemm) grouped_gemm_nt_masked() (in module flashinfer.gemm) grouped_gemm_wgrad() (in module flashinfer.experimental.cake_moe_grouped_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 has_monomoe() (in module flashinfer.fused_moe) hash_topk() (in module flashinfer.fused_moe) head_counts (flashinfer.mla.SparseMLASm120DecodeConfig attribute) hidden_size (flashinfer.fused_moe.CuteDslMoEWrapper attribute) I init_tp_moe_finalize_rmsnorm() (in module flashinfer.comm) init_ulysses_a2a() (in module flashinfer.comm) interleave_for_tma_wgmma_up() (in module flashinfer.fused_moe) 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_buffer_size_sufficient() (flashinfer.comm.MNNVLAllReduceFusionWorkspace method) is_cute_dsl_available() (in module flashinfer.cute_dsl) is_gated_activation() (in module flashinfer.tllm_enums) is_mnnvl_cutedsl_supported() (in module flashinfer.comm) K kernel() (flashinfer.cute_dsl.AddRMSNormFP4QuantKernel method) (flashinfer.cute_dsl.RMSNormFP4QuantKernel method) kv_cache_format (flashinfer.mla.SparseMLASm120DecodeConfig attribute) L layernorm() (in module flashinfer.norm) layernorm_quant() (in module flashinfer.norm) LogitsPipe (class in flashinfer.logits_processor) LogitsProcessor (class in flashinfer.logits_processor) M make_nvfp4_global_scale() (in module flashinfer.quantization) Mapping (class in flashinfer.comm) masked_bmm() (in module flashinfer.gemm) masked_scaled_bmm() (in module flashinfer.gemm) max_num_heads (flashinfer.mla.SparseMLASm120DecodeConfig attribute) max_num_tokens (flashinfer.fused_moe.CuteDslMoEWrapper attribute) (flashinfer.mla.SparseMLASm120DecodeConfig 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) merge_ulysses_output_head_chunk() (in module flashinfer.comm) mhc_post() (in module flashinfer.mhc) mhc_pre_big_fuse() (in module flashinfer.mhc) mhc_pre_big_fuse_with_prenorm() (in module flashinfer.mhc) min_block_table_width() (in module flashinfer) min_p_sampling_from_probs() (in module flashinfer.sampling) min_topk (flashinfer.mla.SparseMLASm120DecodeConfig attribute) minimax_h3_bf16_pre_attention() (in module flashinfer.diffusion_ops) minimax_h3_fc1_swiglu_fp8() (in module flashinfer.diffusion_ops) minimax_h3_fp8_out_proj() (in module flashinfer.diffusion_ops) minimax_h3_fp8_pre_attention() (in module flashinfer.diffusion_ops) minimax_h3_nvfp4_out_proj() (in module flashinfer.diffusion_ops) minimax_h3_nvfp4_pre_attention() (in module flashinfer.diffusion_ops) minimax_h3_sm120_varlen_attention_fp8() (in module flashinfer.diffusion_ops) minimax_h3_sm120_varlen_attention_nvfp4() (in module flashinfer.diffusion_ops) minimax_h3_sm120_varlen_attention_nvfp4_nodelta() (in module flashinfer.diffusion_ops) 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_bf16_fp4() (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_M1_16_K7168_N256_bf16() (in module flashinfer.gemm) mm_M1_16_K7168_N384() (in module flashinfer.gemm) mm_M1_16_K7168_N384_bf16() (in module flashinfer.gemm) mm_M1_16_K7168_N896() (in module flashinfer.gemm) mm_M1_16_K7168_N896_bf16() (in module flashinfer.gemm) mm_mxfp8() (in module flashinfer.gemm) mm_nvfp4_svdquant() (in module flashinfer.gemm) MNNVLAllReduceFusionWorkspace (class in flashinfer.comm) MNNVLCuteDSLAllGatherDualNormWorkspace (class in flashinfer.comm.mnnvl_cutedsl_ag) MNNVLCuteDSLAllGatherNormQuantWorkspace (class in flashinfer.comm.mnnvl_cutedsl_ag) MNNVLCuteDSLAllReduceFusionWorkspace (class in flashinfer.comm.mnnvl_cutedsl_ar) MNNVLCuteDSLMoEFinalizeAllReduceWorkspace (class in flashinfer.comm.mnnvl_cutedsl_moe_finalize) 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_finalize_rmsnorm_allreduce() (in module flashinfer.comm.mnnvl_cutedsl_moe_finalize) moe_gemm_fp8_nt_groupwise() (in module flashinfer.grouped_mm) moe_gemm_mxfp8_nt_groupwise() (in module flashinfer.grouped_mm) MoEActivationPack (class in flashinfer.fused_moe) MoeAlltoAll (class in flashinfer.comm) MoEConfig (class in flashinfer.fused_moe) MoEFinalizeConfig (class in flashinfer.fused_moe) MoELayer (class in flashinfer.fused_moe) MoEWeightPack (class in flashinfer.fused_moe) mono_moe() (in module flashinfer.fused_moe) mpi_barrier() (in module flashinfer.comm.trtllm_mnnvl_ar) msa_proxy_score() (in module flashinfer.msa_ops) msa_proxy_score_fp4() (in module flashinfer.msa_ops) msa_sparse_attention() (in module flashinfer.msa_ops) msa_sparse_decode_attention() (in module flashinfer.msa_ops) msa_topk_select() (in module flashinfer.msa_ops) MSASparseAttentionWorkspace (class in flashinfer.msa_ops) 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_grouped_quantize() (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) NVFP44Over6Config (class in flashinfer.quantization) NVFP44Over6ErrMode (class in flashinfer.quantization) nvfp4_4over6_code() (in module flashinfer.quantization) nvfp4_4over6_from_code() (in module flashinfer.quantization) 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_e4m3_max() (in module flashinfer.quantization) nvfp4_kv_dequantize() (in module flashinfer.quantization) nvfp4_kv_dequantize_paged() (in module flashinfer.quantization) nvfp4_kv_quantize() (in module flashinfer.quantization) nvfp4_quantize() (in module flashinfer.quantization) nvfp4_quantize_append_paged_kv_cache() (in module flashinfer.page) nvfp4_quantize_append_paged_kv_cache_with_slot_mapping() (in module flashinfer.page) nvfp4_quantize_append_paged_mla_kv_cache() (in module flashinfer.page) nvfp4_quantize_append_sparse_mla_cache() (in module flashinfer.mla) nvfp4_quantize_cute_dsl() (in module flashinfer.quantization.kernels.nvfp4_quantize) nvfp4_quantize_pack_sparse_mla_cache() (in module flashinfer.mla) nvfp4_quantize_paged_kv_cache() (in module flashinfer.quantization) nvfp4_quantize_per_token_cute_dsl() (in module flashinfer.quantization.kernels.nvfp4_quantize) nvfp4_quantize_smooth() (in module flashinfer.gemm) P pack_strided_memory() (in module flashinfer.comm) pack_ulysses_qkv_head_chunk() (in module flashinfer.comm) packbits() (in module flashinfer.quantization) packed_fused_kda_decode() (in module flashinfer.kda_decode) packed_kda_decode() (in module flashinfer.kda_decode) padded_seq_len() (in module flashinfer) page_block_size (flashinfer.mla.SparseMLASm120DecodeConfig attribute) page_block_size_is_runtime (flashinfer.mla.SparseMLASm120DecodeConfig attribute) page_block_sizes (flashinfer.mla.SparseMLASm120DecodeConfig attribute) PcieIpcAllGatherLaunchConfig (class in flashinfer.comm) PcieIpcAllGatherVariant (class in flashinfer.comm) PcieIpcAllGatherWorkspace (class in flashinfer.comm) PcieIpcAllReduceWorkspace (class in flashinfer.comm) PcieIpcLaunchConfig (in module flashinfer.comm) PcieIpcReduceScatterLaunchConfig (class in flashinfer.comm) PcieIpcReduceScatterVariant (class in flashinfer.comm) PcieIpcReduceScatterWorkspace (class in flashinfer.comm) PcieIpcVariant (in module flashinfer.comm) per_token_group_quant_8bit() (in module flashinfer.quantization) plan() (flashinfer.attention.BatchAttention method) (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodePagedCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_mla.BatchMLADecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper 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) plan_cake_fmha_request_ordered_paged_decode() (in module flashinfer) plan_cake_vsa() (in module flashinfer.cake_vsa) PODWithPagedKVCacheWrapper (class in flashinfer.pod) precompile_paged_mqa_logits() (in module flashinfer) prepare_all_gather_matmul() (in module flashinfer.comm) prepare_bf16_fp4_weights() (in module flashinfer.gemm) prepare_bgmv_moe() (in module flashinfer.fused_moe) prepare_low_latency_gemm_weights() (in module flashinfer.trtllm_low_latency_gemm) prepare_minimax_h3_fc1_weight_fp8() (in module flashinfer.diffusion_ops) prepare_msa_nvfp4_sparse_decode() (in module flashinfer.msa_ops) prepare_nvfp4_w1_data() (in module flashinfer.fused_moe) prepare_nvfp4_w1_gate_up_data() (in module flashinfer.fused_moe) prepare_nvfp4_w1_gate_up_scales() (in module flashinfer.fused_moe) prepare_nvfp4_w1_scales() (in module flashinfer.fused_moe) prepare_nvfp4_w2_data() (in module flashinfer.fused_moe) prepare_nvfp4_w2_data_k256() (in module flashinfer.fused_moe) prepare_nvfp4_w2_scales() (in module flashinfer.fused_moe) prepare_nvfp4_w2_scales_k256() (in module flashinfer.fused_moe) preprocess_moe_weights_for_sm90_mixed_gemm_humming() (in module flashinfer.fused_moe) prewarm_paged_kv_stride_variant() (flashinfer.attention.BatchAttention method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) prims_ts_bf16_moe() (in module flashinfer.fused_moe) prims_ts_bf16_moe_partitioned() (in module flashinfer.fused_moe) prims_ts_bf16_routed_moe() (in module flashinfer.fused_moe) prims_ts_fp4_block_scale_moe() (in module flashinfer.fused_moe) prims_ts_fp4_block_scale_moe_partitioned() (in module flashinfer.fused_moe) prims_ts_fp4_block_scale_routed_moe() (in module flashinfer.fused_moe) prims_ts_fp8_block_scale_moe() (in module flashinfer.fused_moe) prims_ts_fp8_block_scale_moe_partitioned() (in module flashinfer.fused_moe) prims_ts_fp8_block_scale_routed_moe() (in module flashinfer.fused_moe) prims_ts_fp8_per_tensor_scale_moe() (in module flashinfer.fused_moe) prims_ts_fp8_per_tensor_scale_moe_partitioned() (in module flashinfer.fused_moe) PrimsTsConfig (class in flashinfer.fused_moe) PrimsTsRunner (class in flashinfer.fused_moe) probe_pcie_ipc_rank_topology() (in module flashinfer.comm) probe_ulysses_rank_topology() (in module flashinfer.comm) Q quant_mode (flashinfer.fused_moe.CuteDslMoEWrapper attribute) QuantConfig (class in flashinfer.fused_moe) QuantFormat (class in flashinfer.fused_moe) QuantizationSFLayout (class in flashinfer.comm) quantize_minimax_h3_o_weight_fp8() (in module flashinfer.diffusion_ops) quantize_minimax_h3_o_weight_nvfp4() (in module flashinfer.diffusion_ops) quantize_minimax_h3_qkv_weight_fp8() (in module flashinfer.diffusion_ops) quantize_minimax_h3_qkv_weight_nvfp4() (in module flashinfer.diffusion_ops) quantized_all_reduce() (in module flashinfer.comm) R ragged_block_scaled_bmm() (in module flashinfer.gemm) ragged_bmm() (in module flashinfer.gemm) ragged_scaled_bmm() (in module flashinfer.gemm) recurrent_kda() (in module flashinfer.kda) (in module flashinfer.kda_decode) RecurrentKDAPrefillWorkspace (class in flashinfer.kda_prefill) recv_buffer (flashinfer.comm.UlyssesWorkspace property) refresh_backends() (flashinfer.fused_moe.MoELayer method) reorder_rows_for_gated_act_gemm() (in module flashinfer.fused_moe) replayssm_materialize() (in module flashinfer.mamba) reset_winner() (flashinfer.fused_moe.MoELayer method) reset_workspace_buffer() (flashinfer.cascade.BatchDecodeWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.BatchPrefillWithSharedPrefixPagedKVCacheWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper 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) resolve_nvfp4_4over6() (in module flashinfer.quantization) resolve_pcie_ipc_profile() (in module flashinfer.comm) resolve_ulysses_backend() (in module flashinfer.comm) resolved_backend (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper property) rmsnorm() (in module flashinfer.norm) rmsnorm_fp4quant() (in module flashinfer.cute_dsl) rmsnorm_quant() (in module flashinfer.norm) rmsnorm_sigmoid_gate() (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) RoutingConfig (class in flashinfer.fused_moe) RoutingMethodType (class in flashinfer.fused_moe) run() (flashinfer.attention.BatchAttention method) (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (flashinfer.cascade.MultiLevelCascadeAttentionWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_decode.BatchDecodePagedCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_mla.BatchMLADecodeCuteDSLWrapper method) (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.fused_moe.B12xMoEWrapper method) (flashinfer.fused_moe.CuteDslBf16MoEWrapper method) (flashinfer.fused_moe.CuteDslMoEWrapper method) (flashinfer.fused_moe.CuteDslMxfp8Mxfp4MoEWrapper 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_cake_vsa() (in module flashinfer.cake_vsa) run_mixed_comm() (in module flashinfer.comm.mixed_comm) run_paged() (flashinfer.cute_dsl.attention.wrappers.batch_prefill.BatchPrefillCuteDSLWrapper method) 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) scatter_heads() (flashinfer.comm.UlyssesCommunicator method) scatter_qkv_head_chunk() (flashinfer.comm.UlyssesCommunicator method) segment_packbits() (in module flashinfer.quantization) SegmentGEMMWrapper (class in flashinfer.gemm) selective_state_update() (in module flashinfer.mamba) send_buffer (flashinfer.comm.UlyssesWorkspace property) 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_mxfp8_quantize() (in module flashinfer.gated_act_mxfp8) silu_and_mul_mxfp8_quantize_backward() (in module flashinfer.gated_act_mxfp8) silu_and_mul_nvfp4_quantize() (in module flashinfer.quantization) 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) SparseMLASm120DecodeConfig (class in flashinfer.mla) SparseMLASm120Wrapper (in module flashinfer.mla) split_device_green_ctx() (in module flashinfer.green_ctx) split_device_green_ctx_by_locality_domain() (in module flashinfer.green_ctx) split_device_green_ctx_by_sm_count() (in module flashinfer.green_ctx) ssd_combined_fwd() (in module flashinfer.mamba) stage23_variant_flags() (in module flashinfer.cake_sampling) supported_sparse_mla_sm120_configs() (in module flashinfer.mla) supports_all_gather_gated_residual_next_rmsnorm() (in module flashinfer.comm.mnnvl_cutedsl_ag) supports_all_gather_rmsnorm_fp4_quantize() (in module flashinfer.comm.mnnvl_cutedsl_ag) supports_gated_residual_next_rmsnorm() (in module flashinfer.norm) supports_gated_residual_rmsnorm() (in module flashinfer.norm) supports_moe_finalize_rmsnorm_allreduce() (in module flashinfer.comm.mnnvl_cutedsl_moe_finalize) supports_packed_kv() (in module flashinfer.msa_ops) supports_rmsnorm_sigmoid_gate() (in module flashinfer.norm) supports_tp_moe_finalize_rmsnorm() (in module flashinfer.comm) svdquant_linear() (in module flashinfer.gemm) 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), [1] top_k_mask_logits() (in module flashinfer.sampling) top_k_page_table_transform() (in module flashinfer) top_k_probs_to_slab() (in module flashinfer.cake_sampling) 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.cake_sampling) (in module flashinfer.sampling) top_k_varlen() (in module flashinfer) top_p_renorm_probs() (in module flashinfer.sampling) top_p_sampling_from_probs() (in module flashinfer.sampling) TopK (class in flashinfer.logits_processor) topk_is_runtime (flashinfer.mla.SparseMLASm120DecodeConfig attribute) topks (flashinfer.mla.SparseMLASm120DecodeConfig attribute) TopP (class in flashinfer.logits_processor) tp_moe_finalize_rmsnorm() (in module flashinfer.comm) trtllm_allreduce_fusion() (in module flashinfer.comm) trtllm_batch_context_with_kv_cache() (in module flashinfer.prefill) trtllm_batch_decode_sparse_mla_dsv4() (in module flashinfer.mla) 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_fusion() (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_channel_scale_moe() (in module flashinfer.fused_moe) trtllm_fp8_per_channel_scale_routed_moe() (in module flashinfer.fused_moe) trtllm_fp8_per_tensor_scale_moe() (in module flashinfer.fused_moe) trtllm_fp8_per_tensor_scale_routed_moe() (in module flashinfer.fused_moe) trtllm_gen_routing() (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_prefill_with_kv_cache_mla() (in module flashinfer.mla) trtllm_ragged_attention_deepseek() (in module flashinfer.prefill) TRTLLMAllReduceFusionWorkspace (class in flashinfer.comm) TrtllmGenRoutingResult (class in flashinfer.fused_moe) U ugpu_state_pool() (in module flashinfer.gdn_decode) ulysses_a2a() (in module flashinfer.comm) UlyssesBackendDecision (class in flashinfer.comm) UlyssesBackendError UlyssesCommunicator (class in flashinfer.comm) UlyssesRankTopology (class in flashinfer.comm) UlyssesWorkspace (class in flashinfer.comm) use_cuda_graph (flashinfer.fused_moe.CuteDslMoEWrapper attribute) use_fused_finalize (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) winner_backend (flashinfer.fused_moe.MoELayer property) workspace_size() (flashinfer.attention.BatchPrefillWithCausalBidirectionalRangesWrapper method) (flashinfer.decode.BatchDecodeWithPagedKVCacheWrapper method) (flashinfer.prefill.BatchPrefillWithPagedKVCacheWrapper method) 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)