├── .clang-format ├── .gitignore ├── CMakeLists.txt ├── LICENSE ├── README.md ├── format.sh ├── media └── images │ └── fmha.png ├── performance └── RTX3090 │ ├── decoding_batch_throughput.png │ ├── decoding_seq_throughput.png │ ├── hybrid_throughput.png │ ├── prefill_batch_throughput.png │ └── prefill_seq_throughput.png ├── run_sample.sh ├── src ├── common │ ├── common.h │ ├── cuda_timer.h │ ├── logging.h │ ├── tensor.h │ ├── tester.h │ └── util.h ├── kernel │ ├── flash_attn │ │ ├── fmha.h │ │ ├── fmha │ │ │ ├── alibi.h │ │ │ ├── gemm.h │ │ │ ├── gmem_tile.h │ │ │ ├── kernel_traits.h │ │ │ ├── mask.h │ │ │ ├── smem_tile.h │ │ │ ├── softmax.h │ │ │ └── utils.h │ │ ├── fmha_fprop_kernel_1xN.h │ │ ├── fmha_fwd_hdim128.cu │ │ ├── fmha_fwd_hdim32.cu │ │ ├── fmha_fwd_hdim64.cu │ │ ├── fmha_fwd_launch_template.h │ │ ├── fmha_kernel.h │ │ ├── fmha_utils.h │ │ └── static_switch.h │ └── flash_attn_v2 │ │ ├── alibi.h │ │ ├── block_info.h │ │ ├── flash.h │ │ ├── flash_fwd_hdim128_fp16_sm80.cu │ │ ├── flash_fwd_hdim160_fp16_sm80.cu │ │ ├── flash_fwd_hdim192_fp16_sm80.cu │ │ ├── flash_fwd_hdim224_fp16_sm80.cu │ │ ├── flash_fwd_hdim256_fp16_sm80.cu │ │ ├── flash_fwd_hdim32_fp16_sm80.cu │ │ ├── flash_fwd_hdim64_fp16_sm80.cu │ │ ├── flash_fwd_hdim96_fp16_sm80.cu │ │ ├── flash_fwd_kernel.h │ │ ├── flash_fwd_launch_template.h │ │ ├── kernel_traits.h │ │ ├── kernel_traits_sm90.h │ │ ├── softmax.h │ │ ├── static_switch.h │ │ └── utils.h ├── main.cpp └── ops │ ├── flash_attn.cpp │ ├── flash_attn.h │ ├── flash_attn_v2.cpp │ └── flash_attn_v2.h ├── third_party └── cutlass │ └── include │ ├── cute │ ├── algorithm │ │ ├── axpby.hpp │ │ ├── clear.hpp │ │ ├── copy.hpp │ │ ├── fill.hpp │ │ ├── functional.hpp │ │ ├── gemm.hpp │ │ ├── prefer.hpp │ │ ├── tensor_algorithms.hpp │ │ └── tuple_algorithms.hpp │ ├── arch │ │ ├── cluster_sm90.hpp │ │ ├── copy.hpp │ │ ├── copy_sm75.hpp │ │ ├── copy_sm80.hpp │ │ ├── copy_sm90.hpp │ │ ├── copy_sm90_desc.hpp │ │ ├── copy_sm90_tma.hpp │ │ ├── mma.hpp │ │ ├── mma_sm61.hpp │ │ ├── mma_sm70.hpp │ │ ├── mma_sm75.hpp │ │ ├── mma_sm80.hpp │ │ ├── mma_sm90.hpp │ │ ├── mma_sm90_desc.hpp │ │ ├── mma_sm90_gmma.hpp │ │ └── util.hpp │ ├── atom │ │ ├── copy_atom.hpp │ │ ├── copy_traits.hpp │ │ ├── copy_traits_sm75.hpp │ │ ├── copy_traits_sm80.hpp │ │ ├── copy_traits_sm90.hpp │ │ ├── copy_traits_sm90_tma.hpp │ │ ├── mma_atom.hpp │ │ ├── mma_traits.hpp │ │ ├── mma_traits_sm61.hpp │ │ ├── mma_traits_sm70.hpp │ │ ├── mma_traits_sm75.hpp │ │ ├── mma_traits_sm80.hpp │ │ ├── mma_traits_sm90.hpp │ │ └── mma_traits_sm90_gmma.hpp │ ├── config.hpp │ ├── container │ │ ├── alignment.hpp │ │ ├── array.hpp │ │ ├── array_aligned.hpp │ │ ├── array_subbyte.hpp │ │ ├── bit_field.hpp │ │ ├── cuda_types.hpp │ │ ├── tuple.hpp │ │ └── type_list.hpp │ ├── int_tuple.hpp │ ├── layout.hpp │ ├── numeric │ │ ├── arithmetic_tuple.hpp │ │ ├── bfloat.hpp │ │ ├── complex.hpp │ │ ├── float8.hpp │ │ ├── half.hpp │ │ ├── int.hpp │ │ ├── integer_sequence.hpp │ │ ├── integer_subbyte.hpp │ │ ├── integral_constant.hpp │ │ ├── math.hpp │ │ ├── real.hpp │ │ ├── tfloat.hpp │ │ └── uint128.hpp │ ├── pointer.hpp │ ├── stride.hpp │ ├── swizzle.hpp │ ├── swizzle_layout.hpp │ ├── swizzle_ptr.hpp │ ├── tensor.hpp │ ├── tensor_predicate.hpp │ ├── tile.hpp │ ├── underscore.hpp │ └── util │ │ ├── debug.hpp │ │ ├── print.hpp │ │ └── type_traits.hpp │ └── cutlass │ ├── aligned_buffer.h │ ├── arch │ ├── arch.h │ ├── barrier.h │ ├── cache_operation.h │ ├── memory.h │ ├── memory_sm75.h │ ├── memory_sm80.h │ ├── mma.h │ ├── mma_sm50.h │ ├── mma_sm60.h │ ├── mma_sm61.h │ ├── mma_sm70.h │ ├── mma_sm75.h │ ├── mma_sm80.h │ ├── mma_sm90.h │ ├── mma_sparse_sm80.h │ ├── reg_reconfig.h │ ├── simd.h │ ├── simd_sm60.h │ ├── simd_sm61.h │ ├── wmma.h │ ├── wmma_sm70.h │ ├── wmma_sm72.h │ └── wmma_sm75.h │ ├── array.h │ ├── array_planar_complex.h │ ├── array_subbyte.h │ ├── barrier.h │ ├── bfloat16.h │ ├── blas3.h │ ├── block_striped.h │ ├── cluster_launch.hpp │ ├── complex.h │ ├── constants.h │ ├── conv │ ├── conv2d_problem_size.h │ ├── conv3d_problem_size.h │ ├── convolution.h │ ├── device │ │ ├── direct_convolution.h │ │ ├── implicit_gemm_convolution.h │ │ └── implicit_gemm_convolution_fusion.h │ ├── kernel │ │ ├── default_conv2d.h │ │ ├── default_conv2d_dgrad.h │ │ ├── default_conv2d_fprop.h │ │ ├── default_conv2d_fprop_fusion.h │ │ ├── default_conv2d_fprop_with_broadcast.h │ │ ├── default_conv2d_fprop_with_reduction.h │ │ ├── default_conv2d_group_fprop.h │ │ ├── default_conv2d_wgrad.h │ │ ├── default_conv2d_wgrad_fusion.h │ │ ├── default_conv3d_dgrad.h │ │ ├── default_conv3d_fprop.h │ │ ├── default_conv3d_fprop_fusion.h │ │ ├── default_conv3d_wgrad.h │ │ ├── default_depthwise_fprop.h │ │ ├── direct_convolution.h │ │ ├── implicit_gemm_convolution.h │ │ ├── implicit_gemm_convolution_fusion.h │ │ ├── implicit_gemm_convolution_strided_dgrad.h │ │ └── implicit_gemm_convolution_with_fused_epilogue.h │ ├── thread │ │ └── depthwise_mma.h │ ├── threadblock │ │ ├── conv2d_dgrad_filter_tile_access_iterator_analytic.h │ │ ├── conv2d_dgrad_filter_tile_access_iterator_optimized.h │ │ ├── conv2d_dgrad_output_gradient_tile_access_iterator_analytic.h │ │ ├── conv2d_dgrad_output_gradient_tile_access_iterator_optimized.h │ │ ├── conv2d_fprop_activation_tile_access_iterator_analytic.h │ │ ├── conv2d_fprop_activation_tile_access_iterator_few_channels.h │ │ ├── conv2d_fprop_activation_tile_access_iterator_fixed_channels.h │ │ ├── conv2d_fprop_activation_tile_access_iterator_optimized.h │ │ ├── conv2d_fprop_filter_tile_access_iterator_analytic.h │ │ ├── conv2d_fprop_filter_tile_access_iterator_few_channels.h │ │ ├── conv2d_fprop_filter_tile_access_iterator_fixed_channels.h │ │ ├── conv2d_fprop_filter_tile_access_iterator_optimized.h │ │ ├── conv2d_params.h │ │ ├── conv2d_tile_iterator.h │ │ ├── conv2d_wgrad_activation_tile_access_iterator_analytic.h │ │ ├── conv2d_wgrad_activation_tile_access_iterator_optimized.h │ │ ├── conv2d_wgrad_output_gradient_tile_access_iterator_analytic.h │ │ ├── conv2d_wgrad_output_gradient_tile_access_iterator_optimized.h │ │ ├── conv3d_dgrad_filter_tile_access_iterator_analytic.h │ │ ├── conv3d_dgrad_filter_tile_access_iterator_optimized.h │ │ ├── conv3d_dgrad_output_gradient_tile_access_iterator_analytic.h │ │ ├── conv3d_dgrad_output_gradient_tile_access_iterator_optimized.h │ │ ├── conv3d_fprop_activation_tile_access_iterator_analytic.h │ │ ├── conv3d_fprop_activation_tile_access_iterator_optimized.h │ │ ├── conv3d_fprop_filter_tile_access_iterator_analytic.h │ │ ├── conv3d_fprop_filter_tile_access_iterator_optimized.h │ │ ├── conv3d_params.h │ │ ├── conv3d_wgrad_activation_tile_access_iterator_analytic.h │ │ ├── conv3d_wgrad_activation_tile_access_iterator_optimized.h │ │ ├── conv3d_wgrad_output_gradient_tile_access_iterator_analytic.h │ │ ├── conv3d_wgrad_output_gradient_tile_access_iterator_optimized.h │ │ ├── depthwise_direct_conv_params.h │ │ ├── depthwise_fprop_activation_tile_access_iterator_direct_conv_fixed_stride_dilation.h │ │ ├── depthwise_fprop_activation_tile_access_iterator_direct_conv_optimized.h │ │ ├── depthwise_fprop_direct_conv_multistage.h │ │ ├── depthwise_fprop_filter_tile_access_iterator_direct_conv_optimized.h │ │ ├── depthwise_fprop_pipelined.h │ │ ├── depthwise_mma_base.h │ │ ├── depthwise_mma_core_with_lane_access_size.h │ │ ├── implicit_gemm_fprop_fusion_multistage.h │ │ ├── implicit_gemm_multistage.h │ │ ├── implicit_gemm_pipelined.h │ │ ├── implicit_gemm_wgrad_fusion_multistage.h │ │ ├── predicated_scale_bias_vector_access_iterator.h │ │ ├── predicated_scale_bias_vector_iterator.h │ │ └── threadblock_swizzle.h │ └── warp │ │ ├── mma_depthwise_simt.h │ │ ├── mma_depthwise_simt_tile_iterator.h │ │ └── scale_bias_relu_transform.h │ ├── coord.h │ ├── core_io.h │ ├── cutlass.h │ ├── detail │ └── dependent_false.hpp │ ├── device_kernel.h │ ├── epilogue │ ├── collective │ │ ├── builders │ │ │ └── sm90_builder.inl │ │ ├── collective_builder.hpp │ │ ├── collective_epilogue.hpp │ │ ├── default_epilogue.hpp │ │ ├── detail.hpp │ │ ├── epilogue_tensor_broadcast.hpp │ │ ├── sm70_epilogue_vectorized.hpp │ │ ├── sm90_epilogue_tma_warpspecialized.hpp │ │ └── sm90_epilogue_tma_warpspecialized_bias_elementwise.hpp │ ├── dispatch_policy.hpp │ ├── thread │ │ ├── activation.h │ │ ├── conversion_op.h │ │ ├── detail.hpp │ │ ├── linear_combination.h │ │ ├── linear_combination_bias_elementwise.h │ │ ├── linear_combination_bias_relu.h │ │ ├── linear_combination_clamp.h │ │ ├── linear_combination_dgelu.h │ │ ├── linear_combination_drelu.h │ │ ├── linear_combination_gelu.h │ │ ├── linear_combination_generic.h │ │ ├── linear_combination_hardswish.h │ │ ├── linear_combination_leaky_relu.h │ │ ├── linear_combination_params.h │ │ ├── linear_combination_planar_complex.h │ │ ├── linear_combination_relu.h │ │ ├── linear_combination_relu0.h │ │ ├── linear_combination_residual_block.h │ │ ├── linear_combination_sigmoid.h │ │ ├── linear_combination_silu.h │ │ ├── linear_combination_tensor_broadcast.hpp │ │ ├── linear_combination_with_elementwise.h │ │ ├── reduction_op.h │ │ └── scale_type.h │ ├── threadblock │ │ ├── default_epilogue_complex_tensor_op.h │ │ ├── default_epilogue_complex_tensor_op_blas3.h │ │ ├── default_epilogue_direct_store.h │ │ ├── default_epilogue_planar_complex.h │ │ ├── default_epilogue_simt.h │ │ ├── default_epilogue_tensor_op.h │ │ ├── default_epilogue_tensor_op_blas3.h │ │ ├── default_epilogue_tensor_op_row_broadcast.h │ │ ├── default_epilogue_volta_tensor_op.h │ │ ├── default_epilogue_with_broadcast.h │ │ ├── default_epilogue_with_reduction.h │ │ ├── default_epilogue_wmma_tensor_op.h │ │ ├── default_thread_map_simt.h │ │ ├── default_thread_map_tensor_op.h │ │ ├── default_thread_map_volta_tensor_op.h │ │ ├── default_thread_map_wmma_tensor_op.h │ │ ├── direct_store_epilogue_iterator.h │ │ ├── epilogue.h │ │ ├── epilogue_base.h │ │ ├── epilogue_base_streamk.h │ │ ├── epilogue_depthwise.h │ │ ├── epilogue_direct_store.h │ │ ├── epilogue_gemm_k_reduction.h │ │ ├── epilogue_planar_complex.h │ │ ├── epilogue_smem_accumulator.h │ │ ├── epilogue_streamk_with_broadcast.h │ │ ├── epilogue_visitor_with_softmax.h │ │ ├── epilogue_with_broadcast.h │ │ ├── epilogue_with_reduction.h │ │ ├── epilogue_with_visitor.h │ │ ├── epilogue_workspace.h │ │ ├── interleaved_epilogue.h │ │ ├── output_iterator_parameter.h │ │ ├── output_tile_thread_map.h │ │ ├── predicated_tile_iterator.h │ │ ├── predicated_tile_iterator_affine.h │ │ ├── predicated_tile_iterator_affine_layout_params.h │ │ ├── predicated_tile_iterator_blas3.h │ │ ├── predicated_tile_iterator_direct_conv.h │ │ ├── predicated_tile_iterator_params.h │ │ ├── predicated_tile_iterator_predicates.h │ │ ├── predicated_tile_iterator_row_broadcast.h │ │ ├── predicated_tile_iterator_strided_dgrad.h │ │ ├── shared_load_iterator.h │ │ ├── shared_load_iterator_mixed.h │ │ └── shared_load_iterator_pitch_liner.h │ └── warp │ │ ├── fragment_iterator_complex_tensor_op.h │ │ ├── fragment_iterator_gaussian_complex_tensor_op.h │ │ ├── fragment_iterator_simt.h │ │ ├── fragment_iterator_tensor_op.h │ │ ├── fragment_iterator_volta_tensor_op.h │ │ ├── fragment_iterator_wmma_tensor_op.h │ │ ├── simt_policy.h │ │ ├── tensor_op_policy.h │ │ ├── tile_iterator_simt.h │ │ ├── tile_iterator_tensor_op.h │ │ ├── tile_iterator_tensor_op_mixed.h │ │ ├── tile_iterator_volta_tensor_op.h │ │ ├── tile_iterator_wmma_tensor_op.h │ │ ├── volta_tensor_op_policy.h │ │ └── wmma_tensor_op_policy.h │ ├── fast_math.h │ ├── float8.h │ ├── floating_point_nvrtc.h │ ├── functional.h │ ├── gemm │ ├── collective │ │ ├── builders │ │ │ └── sm90_gmma_builder.inl │ │ ├── collective_builder.hpp │ │ ├── collective_mma.hpp │ │ ├── sm70_mma_twostage.hpp │ │ ├── sm80_mma_multistage.hpp │ │ ├── sm90_mma_multistage_gmma_ss.hpp │ │ ├── sm90_mma_tma_gmma_rs_warpspecialized.hpp │ │ ├── sm90_mma_tma_gmma_ss.hpp │ │ └── sm90_mma_tma_gmma_ss_warpspecialized.hpp │ ├── device │ │ ├── base_grouped.h │ │ ├── default_gemm_configuration.h │ │ ├── ell_gemm.h │ │ ├── gemm.h │ │ ├── gemm_array.h │ │ ├── gemm_batched.h │ │ ├── gemm_complex.h │ │ ├── gemm_grouped.h │ │ ├── gemm_layernorm_mainloop_fusion.h │ │ ├── gemm_sparse.h │ │ ├── gemm_sparse_row_broadcast.h │ │ ├── gemm_splitk_parallel.h │ │ ├── gemm_universal.h │ │ ├── gemm_universal_adapter.h │ │ ├── gemm_universal_base.h │ │ ├── gemm_universal_streamk_with_broadcast.h │ │ ├── gemm_universal_with_broadcast.h │ │ ├── gemm_with_k_reduction.h │ │ ├── gemv.h │ │ ├── rank_2k.h │ │ ├── rank_2k_grouped.h │ │ ├── rank_k.h │ │ ├── symm.h │ │ └── trmm.h │ ├── dispatch_policy.hpp │ ├── gemm.h │ ├── kernel │ │ ├── default_ell_gemm.h │ │ ├── default_gemm.h │ │ ├── default_gemm_complex.h │ │ ├── default_gemm_grouped.h │ │ ├── default_gemm_grouped_softmax_mainloop_fusion.h │ │ ├── default_gemm_layernorm_mainloop_fusion.h │ │ ├── default_gemm_planar_complex_universal.h │ │ ├── default_gemm_sparse.h │ │ ├── default_gemm_sparse_row_broadcast.h │ │ ├── default_gemm_splitk_parallel.h │ │ ├── default_gemm_streamk_with_broadcast.h │ │ ├── default_gemm_universal.h │ │ ├── default_gemm_with_broadcast.h │ │ ├── default_gemm_with_k_reduction.h │ │ ├── default_gemm_with_reduction.h │ │ ├── default_gemv.h │ │ ├── default_rank_2k.h │ │ ├── default_rank_2k_complex.h │ │ ├── default_rank_2k_grouped.h │ │ ├── default_rank_2k_universal.h │ │ ├── default_rank_k.h │ │ ├── default_rank_k_complex.h │ │ ├── default_rank_k_universal.h │ │ ├── default_symm.h │ │ ├── default_symm_complex.h │ │ ├── default_symm_universal.h │ │ ├── default_trmm.h │ │ ├── default_trmm_complex.h │ │ ├── default_trmm_universal.h │ │ ├── ell_gemm.h │ │ ├── gemm.h │ │ ├── gemm_array.h │ │ ├── gemm_batched.h │ │ ├── gemm_grouped.h │ │ ├── gemm_grouped_problem_visitor.h │ │ ├── gemm_grouped_softmax_mainloop_fusion.h │ │ ├── gemm_layernorm_mainloop_fusion.h │ │ ├── gemm_params.h │ │ ├── gemm_pipelined.h │ │ ├── gemm_planar_complex.h │ │ ├── gemm_planar_complex_array.h │ │ ├── gemm_splitk_parallel.h │ │ ├── gemm_streamk_with_fused_epilogue.h │ │ ├── gemm_transpose_operands.h │ │ ├── gemm_universal.h │ │ ├── gemm_universal.hpp │ │ ├── gemm_universal_streamk.h │ │ ├── gemm_with_fused_epilogue.h │ │ ├── gemm_with_k_reduction.h │ │ ├── gemv.h │ │ ├── gemv_batched_strided.h │ │ ├── grouped_problem_visitor.h │ │ ├── params_universal_base.h │ │ ├── rank_2k_grouped.h │ │ ├── rank_2k_grouped_problem_visitor.h │ │ ├── rank_2k_transpose_operands.h │ │ ├── rank_2k_universal.h │ │ ├── rank_k_universal.h │ │ ├── sm70_gemm.hpp │ │ ├── sm90_gemm_tma.hpp │ │ ├── sm90_gemm_tma_warpspecialized.hpp │ │ ├── sm90_gemm_tma_warpspecialized_cooperative.hpp │ │ ├── sm90_gemm_tma_warpspecialized_pingpong.hpp │ │ ├── sm90_tile_scheduler.hpp │ │ ├── sparse_gemm.h │ │ ├── sparse_gemm_row_broadcast.h │ │ ├── symm_universal.h │ │ └── trmm_universal.h │ ├── thread │ │ ├── mma.h │ │ ├── mma_sm50.h │ │ ├── mma_sm60.h │ │ └── mma_sm61.h │ ├── threadblock │ │ ├── default_ell_mma.h │ │ ├── default_gemv_core.h │ │ ├── default_mma.h │ │ ├── default_mma_core.h │ │ ├── default_mma_core_simt.h │ │ ├── default_mma_core_sm70.h │ │ ├── default_mma_core_sm75.h │ │ ├── default_mma_core_sm80.h │ │ ├── default_mma_core_sparse_sm80.h │ │ ├── default_mma_core_with_access_size.h │ │ ├── default_mma_core_with_reduction.h │ │ ├── default_mma_core_wmma.h │ │ ├── default_mma_layernorm_mainloop_fusion.h │ │ ├── default_mma_planar_complex_multistage.h │ │ ├── default_mma_planar_complex_pipelined.h │ │ ├── default_mma_softmax_mainloop_fusion.h │ │ ├── default_mma_with_reduction.h │ │ ├── default_multistage_mma_complex.h │ │ ├── default_multistage_mma_complex_core.h │ │ ├── default_multistage_mma_complex_core_sm80.h │ │ ├── default_multistage_trmm_complex.h │ │ ├── default_sparse_mma.h │ │ ├── default_trmm.h │ │ ├── ell_mma_multistage.h │ │ ├── ell_mma_pipelined.h │ │ ├── gemv.h │ │ ├── index_remat.h │ │ ├── mma_base.h │ │ ├── mma_blas3_multistage.h │ │ ├── mma_layernorm_mainloop_fusion_multistage.h │ │ ├── mma_multistage.h │ │ ├── mma_pipelined.h │ │ ├── mma_planar_complex_base.h │ │ ├── mma_planar_complex_multistage.h │ │ ├── mma_planar_complex_pipelined.h │ │ ├── mma_singlestage.h │ │ ├── mma_softmax_mainloop_fusion_multistage.h │ │ ├── mma_sparse_base.h │ │ ├── mma_sparse_multistage.h │ │ ├── mma_with_reduction_multistage.h │ │ ├── threadblock_swizzle.h │ │ └── threadblock_swizzle_streamk.h │ └── warp │ │ ├── default_mma_complex_tensor_op.h │ │ ├── default_mma_sparse_tensor_op.h │ │ ├── default_mma_tensor_op.h │ │ ├── default_mma_tensor_op_sm80.h │ │ ├── default_mma_with_reduction_tensor_op.h │ │ ├── default_mma_wmma_tensor_op.h │ │ ├── layernorm_scale_bias_transform.h │ │ ├── mma.h │ │ ├── mma_complex_tensor_op.h │ │ ├── mma_complex_tensor_op_fast_f32.h │ │ ├── mma_complex_tensor_op_tile_iterator_sm80.h │ │ ├── mma_gaussian_complex_tensor_op.h │ │ ├── mma_gaussian_complex_tensor_op_tile_iterator_sm80.h │ │ ├── mma_planar_complex.h │ │ ├── mma_simt.h │ │ ├── mma_simt_policy.h │ │ ├── mma_simt_tile_iterator.h │ │ ├── mma_sparse_tensor_op.h │ │ ├── mma_tensor_op.h │ │ ├── mma_tensor_op_fast_f32.h │ │ ├── mma_tensor_op_fragment_iterator.h │ │ ├── mma_tensor_op_policy.h │ │ ├── mma_tensor_op_sm70.h │ │ ├── mma_tensor_op_tile_access_iterator.h │ │ ├── mma_tensor_op_tile_iterator.h │ │ ├── mma_tensor_op_tile_iterator_sm70.h │ │ ├── mma_tensor_op_tile_iterator_sm80.h │ │ ├── mma_tensor_op_tile_iterator_sparse.h │ │ ├── mma_tensor_op_tile_iterator_wmma.h │ │ ├── mma_tensor_op_wmma.h │ │ ├── mma_with_reduction_tensor_op.h │ │ ├── scale_bias_tile_iterator.h │ │ ├── softmax_scale_bias_transform.h │ │ └── tile_iterator_planar_complex.h │ ├── half.h │ ├── integer_subbyte.h │ ├── kernel_hardware_info.hpp │ ├── kernel_launch.h │ ├── layout │ ├── layout.h │ ├── matrix.h │ ├── permute.h │ ├── pitch_linear.h │ ├── tensor.h │ ├── tensor_op_multiplicand_sm70.h │ ├── tensor_op_multiplicand_sm75.h │ ├── tensor_op_multiplicand_sm80.h │ └── vector.h │ ├── matrix.h │ ├── matrix_coord.h │ ├── matrix_shape.h │ ├── numeric_conversion.h │ ├── numeric_types.h │ ├── pipeline │ ├── pipeline.hpp │ └── sm90_pipeline.hpp │ ├── pitch_linear_coord.h │ ├── platform │ └── platform.h │ ├── predicate_vector.h │ ├── quaternion.h │ ├── real.h │ ├── reduction │ ├── device │ │ ├── reduce_split_k.h │ │ ├── tensor_reduce.h │ │ ├── tensor_reduce_affine_contiguous.h │ │ └── tensor_reduce_affine_strided.h │ ├── kernel │ │ ├── reduce_softmax_final.h │ │ ├── reduce_split_k.h │ │ ├── tensor_reduce_affine_contiguous.h │ │ └── tensor_reduce_affine_strided.h │ ├── thread │ │ ├── reduce.h │ │ └── reduction_operators.h │ └── threadblock_swizzle.h │ ├── relatively_equal.h │ ├── semaphore.h │ ├── subbyte_reference.h │ ├── tensor_coord.h │ ├── tensor_ref.h │ ├── tensor_ref_planar_complex.h │ ├── tensor_view.h │ ├── tensor_view_planar_complex.h │ ├── tfloat32.h │ ├── thread │ └── matrix.h │ ├── trace.h │ ├── transform │ ├── collective │ │ └── sm90_wgmma_transpose.hpp │ ├── pitch_linear_thread_map.h │ ├── thread │ │ ├── transpose.h │ │ └── unary_op.h │ ├── threadblock │ │ ├── ell_iterator.h │ │ ├── ell_predicated_tile_access_iterator.h │ │ ├── ell_predicated_tile_iterator.h │ │ ├── predicated_scale_bias_vector_access_iterator.h │ │ ├── predicated_scale_bias_vector_iterator.h │ │ ├── predicated_tile_access_iterator.h │ │ ├── predicated_tile_access_iterator_2dthreadtile.h │ │ ├── predicated_tile_access_iterator_params.h │ │ ├── predicated_tile_access_iterator_triangular_matrix.h │ │ ├── predicated_tile_iterator.h │ │ ├── predicated_tile_iterator_2dthreadtile.h │ │ ├── predicated_tile_iterator_triangular_matrix.h │ │ ├── predicated_vector_access_iterator.h │ │ ├── regular_scale_bias_vector_access_iterator.h │ │ ├── regular_tile_access_iterator.h │ │ ├── regular_tile_access_iterator_pitch_linear.h │ │ ├── regular_tile_access_iterator_pitch_linear_direct_conv.h │ │ ├── regular_tile_access_iterator_tensor_op.h │ │ ├── regular_tile_access_iterator_tensor_op_sm80.h │ │ ├── regular_tile_iterator.h │ │ ├── regular_tile_iterator_pitch_linear.h │ │ ├── regular_tile_iterator_pitch_linear_2dthreadtile.h │ │ ├── regular_tile_iterator_tensor_op.h │ │ ├── regular_tile_iterator_tensor_op_sm70.h │ │ └── vector_iterator.h │ └── warp │ │ └── vector_fragment_iterator.h │ ├── uint128.h │ └── wmma_array.h └── tools └── performance ├── performance.py └── performance.sh /.clang-format: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/.clang-format -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/.gitignore -------------------------------------------------------------------------------- /CMakeLists.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/CMakeLists.txt -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/README.md -------------------------------------------------------------------------------- /format.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/format.sh -------------------------------------------------------------------------------- /media/images/fmha.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/media/images/fmha.png -------------------------------------------------------------------------------- /performance/RTX3090/decoding_batch_throughput.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/performance/RTX3090/decoding_batch_throughput.png -------------------------------------------------------------------------------- /performance/RTX3090/decoding_seq_throughput.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/performance/RTX3090/decoding_seq_throughput.png -------------------------------------------------------------------------------- /performance/RTX3090/hybrid_throughput.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/performance/RTX3090/hybrid_throughput.png -------------------------------------------------------------------------------- /performance/RTX3090/prefill_batch_throughput.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/performance/RTX3090/prefill_batch_throughput.png -------------------------------------------------------------------------------- /performance/RTX3090/prefill_seq_throughput.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/performance/RTX3090/prefill_seq_throughput.png -------------------------------------------------------------------------------- /run_sample.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/run_sample.sh -------------------------------------------------------------------------------- /src/common/common.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/common.h -------------------------------------------------------------------------------- /src/common/cuda_timer.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/cuda_timer.h -------------------------------------------------------------------------------- /src/common/logging.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/logging.h -------------------------------------------------------------------------------- /src/common/tensor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/tensor.h -------------------------------------------------------------------------------- /src/common/tester.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/tester.h -------------------------------------------------------------------------------- /src/common/util.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/common/util.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/alibi.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/alibi.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/gemm.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/gmem_tile.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/gmem_tile.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/kernel_traits.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/kernel_traits.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/mask.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/mask.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/smem_tile.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/smem_tile.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/softmax.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/softmax.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha/utils.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha/utils.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_fprop_kernel_1xN.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_fprop_kernel_1xN.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_fwd_hdim128.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_fwd_hdim128.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_fwd_hdim32.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_fwd_hdim32.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_fwd_hdim64.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_fwd_hdim64.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_fwd_launch_template.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_fwd_launch_template.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_kernel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_kernel.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/fmha_utils.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/fmha_utils.h -------------------------------------------------------------------------------- /src/kernel/flash_attn/static_switch.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn/static_switch.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/alibi.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/alibi.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/block_info.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/block_info.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim128_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim128_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim160_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim160_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim192_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim192_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim224_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim224_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim256_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim256_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim32_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim32_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim64_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim64_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_hdim96_fp16_sm80.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_hdim96_fp16_sm80.cu -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_kernel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_kernel.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/flash_fwd_launch_template.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/flash_fwd_launch_template.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/kernel_traits.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/kernel_traits.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/kernel_traits_sm90.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/kernel_traits_sm90.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/softmax.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/softmax.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/static_switch.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/static_switch.h -------------------------------------------------------------------------------- /src/kernel/flash_attn_v2/utils.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/kernel/flash_attn_v2/utils.h -------------------------------------------------------------------------------- /src/main.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/main.cpp -------------------------------------------------------------------------------- /src/ops/flash_attn.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/ops/flash_attn.cpp -------------------------------------------------------------------------------- /src/ops/flash_attn.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/ops/flash_attn.h -------------------------------------------------------------------------------- /src/ops/flash_attn_v2.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/ops/flash_attn_v2.cpp -------------------------------------------------------------------------------- /src/ops/flash_attn_v2.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/src/ops/flash_attn_v2.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/axpby.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/axpby.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/clear.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/clear.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/copy.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/copy.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/fill.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/fill.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/functional.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/functional.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/gemm.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/gemm.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/prefer.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/prefer.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/tensor_algorithms.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/tensor_algorithms.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/algorithm/tuple_algorithms.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/algorithm/tuple_algorithms.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/cluster_sm90.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/cluster_sm90.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy_sm75.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy_sm75.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy_sm80.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy_sm80.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy_sm90.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy_sm90.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy_sm90_desc.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy_sm90_desc.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/copy_sm90_tma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/copy_sm90_tma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm61.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm61.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm70.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm70.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm75.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm75.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm80.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm80.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm90.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm90.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm90_desc.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm90_desc.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/mma_sm90_gmma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/mma_sm90_gmma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/arch/util.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/arch/util.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_atom.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_atom.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_traits.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_traits.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_traits_sm75.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_traits_sm75.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_traits_sm80.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_traits_sm80.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_traits_sm90.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_traits_sm90.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/copy_traits_sm90_tma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/copy_traits_sm90_tma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_atom.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_atom.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm61.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm61.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm70.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm70.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm75.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm75.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm80.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm80.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm90.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm90.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/atom/mma_traits_sm90_gmma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/atom/mma_traits_sm90_gmma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/config.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/config.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/alignment.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/alignment.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/array.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/array.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/array_aligned.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/array_aligned.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/array_subbyte.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/array_subbyte.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/bit_field.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/bit_field.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/cuda_types.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/cuda_types.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/tuple.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/tuple.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/container/type_list.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/container/type_list.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/int_tuple.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/int_tuple.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/layout.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/layout.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/arithmetic_tuple.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/arithmetic_tuple.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/bfloat.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/bfloat.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/complex.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/complex.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/float8.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/float8.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/half.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/half.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/int.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/int.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/integer_sequence.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/integer_sequence.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/integer_subbyte.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/integer_subbyte.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/integral_constant.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/integral_constant.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/math.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/math.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/real.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/real.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/tfloat.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/tfloat.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/numeric/uint128.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/numeric/uint128.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/pointer.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/pointer.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/stride.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/stride.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/swizzle.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/swizzle.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/swizzle_layout.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/swizzle_layout.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/swizzle_ptr.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/swizzle_ptr.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/tensor.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/tensor.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/tensor_predicate.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/tensor_predicate.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/tile.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/tile.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/underscore.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/underscore.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/util/debug.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/util/debug.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/util/print.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/util/print.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cute/util/type_traits.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cute/util/type_traits.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/aligned_buffer.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/aligned_buffer.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/arch.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/arch.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/barrier.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/barrier.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/cache_operation.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/cache_operation.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/memory.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/memory.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/memory_sm75.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/memory_sm75.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/memory_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/memory_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm50.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm50.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm60.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm60.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm61.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm61.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm75.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm75.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sm90.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sm90.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/mma_sparse_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/mma_sparse_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/reg_reconfig.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/reg_reconfig.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/simd.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/simd.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/simd_sm60.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/simd_sm60.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/simd_sm61.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/simd_sm61.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/wmma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/wmma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/wmma_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/wmma_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/wmma_sm72.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/wmma_sm72.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/arch/wmma_sm75.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/arch/wmma_sm75.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/array.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/array.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/array_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/array_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/array_subbyte.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/array_subbyte.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/barrier.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/barrier.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/bfloat16.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/bfloat16.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/blas3.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/blas3.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/block_striped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/block_striped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/cluster_launch.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/cluster_launch.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/constants.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/constants.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/conv2d_problem_size.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/conv2d_problem_size.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/conv3d_problem_size.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/conv3d_problem_size.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/convolution.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/convolution.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/device/direct_convolution.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/device/direct_convolution.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/device/implicit_gemm_convolution.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/device/implicit_gemm_convolution.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/device/implicit_gemm_convolution_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/device/implicit_gemm_convolution_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_dgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_dgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_fprop_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_group_fprop.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_group_fprop.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_wgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_wgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_wgrad_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv2d_wgrad_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_dgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_dgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_fprop.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_fprop.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_fprop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_fprop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_wgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_conv3d_wgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/default_depthwise_fprop.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/default_depthwise_fprop.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/direct_convolution.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/direct_convolution.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_strided_dgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_strided_dgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_with_fused_epilogue.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/kernel/implicit_gemm_convolution_with_fused_epilogue.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/thread/depthwise_mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/thread/depthwise_mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_filter_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_filter_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_filter_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_filter_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_output_gradient_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_output_gradient_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_output_gradient_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_dgrad_output_gradient_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_few_channels.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_few_channels.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_fixed_channels.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_fixed_channels.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_few_channels.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_few_channels.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_fixed_channels.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_fixed_channels.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_filter_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_activation_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_activation_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_activation_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_activation_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_output_gradient_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_output_gradient_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_output_gradient_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv2d_wgrad_output_gradient_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_filter_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_filter_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_filter_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_filter_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_output_gradient_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_output_gradient_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_output_gradient_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_dgrad_output_gradient_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_activation_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_activation_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_activation_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_activation_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_filter_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_filter_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_filter_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_filter_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_activation_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_activation_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_activation_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_activation_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_output_gradient_tile_access_iterator_analytic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_output_gradient_tile_access_iterator_analytic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_output_gradient_tile_access_iterator_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/conv3d_wgrad_output_gradient_tile_access_iterator_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_direct_conv_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_direct_conv_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_activation_tile_access_iterator_direct_conv_fixed_stride_dilation.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_activation_tile_access_iterator_direct_conv_fixed_stride_dilation.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_activation_tile_access_iterator_direct_conv_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_activation_tile_access_iterator_direct_conv_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_direct_conv_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_direct_conv_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_filter_tile_access_iterator_direct_conv_optimized.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_filter_tile_access_iterator_direct_conv_optimized.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_fprop_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_mma_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_mma_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/depthwise_mma_core_with_lane_access_size.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/depthwise_mma_core_with_lane_access_size.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_fprop_fusion_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_fprop_fusion_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_wgrad_fusion_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/implicit_gemm_wgrad_fusion_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/threadblock/threadblock_swizzle.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/threadblock/threadblock_swizzle.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/warp/mma_depthwise_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/warp/mma_depthwise_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/warp/mma_depthwise_simt_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/warp/mma_depthwise_simt_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/conv/warp/scale_bias_relu_transform.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/conv/warp/scale_bias_relu_transform.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/coord.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/coord.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/core_io.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/core_io.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/cutlass.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/cutlass.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/detail/dependent_false.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/detail/dependent_false.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/device_kernel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/device_kernel.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/builders/sm90_builder.inl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/builders/sm90_builder.inl -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/collective_builder.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/collective_builder.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/collective_epilogue.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/collective_epilogue.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/default_epilogue.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/default_epilogue.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/detail.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/detail.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/epilogue_tensor_broadcast.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/epilogue_tensor_broadcast.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/sm70_epilogue_vectorized.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/sm70_epilogue_vectorized.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/sm90_epilogue_tma_warpspecialized.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/sm90_epilogue_tma_warpspecialized.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/collective/sm90_epilogue_tma_warpspecialized_bias_elementwise.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/collective/sm90_epilogue_tma_warpspecialized_bias_elementwise.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/dispatch_policy.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/dispatch_policy.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/activation.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/activation.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/conversion_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/conversion_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/detail.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/detail.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_bias_elementwise.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_bias_elementwise.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_bias_relu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_bias_relu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_clamp.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_clamp.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_dgelu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_dgelu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_drelu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_drelu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_gelu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_gelu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_generic.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_generic.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_hardswish.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_hardswish.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_leaky_relu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_leaky_relu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_relu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_relu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_relu0.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_relu0.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_residual_block.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_residual_block.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_sigmoid.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_sigmoid.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_silu.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_silu.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_tensor_broadcast.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_tensor_broadcast.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_with_elementwise.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/linear_combination_with_elementwise.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/reduction_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/reduction_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/thread/scale_type.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/thread/scale_type.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_complex_tensor_op_blas3.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_complex_tensor_op_blas3.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_direct_store.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_direct_store.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op_blas3.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op_blas3.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op_row_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_tensor_op_row_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_volta_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_volta_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_wmma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_epilogue_wmma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_volta_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_volta_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_wmma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/default_thread_map_wmma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/direct_store_epilogue_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/direct_store_epilogue_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_base_streamk.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_base_streamk.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_depthwise.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_depthwise.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_direct_store.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_direct_store.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_gemm_k_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_gemm_k_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_smem_accumulator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_smem_accumulator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_streamk_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_streamk_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_visitor_with_softmax.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_visitor_with_softmax.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_visitor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_with_visitor.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_workspace.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/epilogue_workspace.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/interleaved_epilogue.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/interleaved_epilogue.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/output_iterator_parameter.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/output_iterator_parameter.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/output_tile_thread_map.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/output_tile_thread_map.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_affine.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_affine.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_affine_layout_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_affine_layout_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_blas3.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_blas3.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_direct_conv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_direct_conv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_predicates.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_predicates.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_row_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_row_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_strided_dgrad.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_strided_dgrad.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator_mixed.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator_mixed.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator_pitch_liner.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/threadblock/shared_load_iterator_pitch_liner.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_gaussian_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_gaussian_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_volta_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_volta_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_wmma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/fragment_iterator_wmma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/simt_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/simt_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tensor_op_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tensor_op_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op_mixed.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op_mixed.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_volta_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_volta_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_wmma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/tile_iterator_wmma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/volta_tensor_op_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/volta_tensor_op_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/epilogue/warp/wmma_tensor_op_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/epilogue/warp/wmma_tensor_op_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/fast_math.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/fast_math.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/float8.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/float8.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/floating_point_nvrtc.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/floating_point_nvrtc.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/functional.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/functional.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/builders/sm90_gmma_builder.inl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/builders/sm90_gmma_builder.inl -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/collective_builder.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/collective_builder.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/collective_mma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/collective_mma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm70_mma_twostage.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm70_mma_twostage.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm80_mma_multistage.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm80_mma_multistage.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_multistage_gmma_ss.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_multistage_gmma_ss.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_rs_warpspecialized.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_rs_warpspecialized.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_ss.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_ss.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_ss_warpspecialized.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/collective/sm90_mma_tma_gmma_ss_warpspecialized.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/base_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/base_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/default_gemm_configuration.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/default_gemm_configuration.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/ell_gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/ell_gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_array.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_array.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_batched.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_batched.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_layernorm_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_layernorm_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_sparse.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_sparse.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_sparse_row_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_sparse_row_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_splitk_parallel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_splitk_parallel.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_universal_adapter.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_universal_adapter.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_universal_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_universal_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_universal_streamk_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_universal_streamk_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_universal_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_universal_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemm_with_k_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemm_with_k_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/gemv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/gemv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/rank_2k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/rank_2k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/rank_2k_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/rank_2k_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/rank_k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/rank_k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/symm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/symm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/device/trmm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/device/trmm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/dispatch_policy.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/dispatch_policy.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_ell_gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_ell_gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_grouped_softmax_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_grouped_softmax_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_layernorm_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_layernorm_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_planar_complex_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_planar_complex_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_sparse.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_sparse.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_sparse_row_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_sparse_row_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_splitk_parallel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_splitk_parallel.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_streamk_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_streamk_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_k_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_k_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemm_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_gemv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_gemv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_2k_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_rank_k_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_symm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_symm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_symm_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_symm_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_symm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_symm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_trmm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_trmm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_trmm_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_trmm_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/default_trmm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/default_trmm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/ell_gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/ell_gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_array.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_array.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_batched.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_batched.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped_problem_visitor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped_problem_visitor.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped_softmax_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_grouped_softmax_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_layernorm_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_layernorm_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_planar_complex_array.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_planar_complex_array.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_splitk_parallel.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_splitk_parallel.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_streamk_with_fused_epilogue.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_streamk_with_fused_epilogue.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_transpose_operands.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_transpose_operands.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal_streamk.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_universal_streamk.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_with_fused_epilogue.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_with_fused_epilogue.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemm_with_k_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemm_with_k_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/gemv_batched_strided.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/gemv_batched_strided.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/grouped_problem_visitor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/grouped_problem_visitor.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/params_universal_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/params_universal_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_grouped.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_grouped.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_grouped_problem_visitor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_grouped_problem_visitor.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_transpose_operands.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_transpose_operands.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/rank_2k_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/rank_k_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/rank_k_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm70_gemm.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm70_gemm.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_cooperative.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_cooperative.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_pingpong.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_pingpong.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sparse_gemm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sparse_gemm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/sparse_gemm_row_broadcast.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/sparse_gemm_row_broadcast.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/symm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/symm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/kernel/trmm_universal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/kernel/trmm_universal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/thread/mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/thread/mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/thread/mma_sm50.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/thread/mma_sm50.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/thread/mma_sm60.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/thread/mma_sm60.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/thread/mma_sm61.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/thread/mma_sm61.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_ell_mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_ell_mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_gemv_core.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_gemv_core.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm75.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm75.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sparse_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_sparse_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_with_access_size.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_with_access_size.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_wmma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_core_wmma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_layernorm_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_layernorm_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_planar_complex_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_planar_complex_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_planar_complex_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_planar_complex_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_softmax_mainloop_fusion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_softmax_mainloop_fusion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_with_reduction.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_mma_with_reduction.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex_core.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex_core.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex_core_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_mma_complex_core_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_trmm_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_multistage_trmm_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_sparse_mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_sparse_mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/default_trmm.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/default_trmm.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/ell_mma_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/ell_mma_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/ell_mma_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/ell_mma_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/gemv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/gemv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/index_remat.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/index_remat.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_blas3_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_blas3_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_layernorm_mainloop_fusion_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_layernorm_mainloop_fusion_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_pipelined.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_planar_complex_pipelined.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_singlestage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_singlestage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_softmax_mainloop_fusion_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_softmax_mainloop_fusion_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_sparse_base.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_sparse_base.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_sparse_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_sparse_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/mma_with_reduction_multistage.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/mma_with_reduction_multistage.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/threadblock_swizzle.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/threadblock_swizzle.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/threadblock/threadblock_swizzle_streamk.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/threadblock/threadblock_swizzle_streamk.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_sparse_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_sparse_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_tensor_op_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_tensor_op_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_with_reduction_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_with_reduction_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/default_mma_wmma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/default_mma_wmma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/layernorm_scale_bias_transform.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/layernorm_scale_bias_transform.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_fast_f32.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_fast_f32.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_tile_iterator_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_tile_iterator_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_gaussian_complex_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_gaussian_complex_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_gaussian_complex_tensor_op_tile_iterator_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_gaussian_complex_tensor_op_tile_iterator_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_simt.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_simt.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_simt_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_simt_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_simt_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_simt_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_sparse_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_sparse_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_fast_f32.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_fast_f32.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_fragment_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_fragment_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_policy.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_policy.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sparse.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sparse.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_wmma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_wmma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_wmma.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_tensor_op_wmma.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/mma_with_reduction_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/mma_with_reduction_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/scale_bias_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/scale_bias_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/softmax_scale_bias_transform.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/softmax_scale_bias_transform.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/gemm/warp/tile_iterator_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/gemm/warp/tile_iterator_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/half.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/half.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/integer_subbyte.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/integer_subbyte.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/kernel_hardware_info.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/kernel_hardware_info.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/kernel_launch.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/kernel_launch.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/layout.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/layout.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/matrix.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/matrix.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/permute.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/permute.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/pitch_linear.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/pitch_linear.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/tensor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/tensor.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm75.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm75.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/tensor_op_multiplicand_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/layout/vector.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/layout/vector.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/matrix.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/matrix.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/matrix_coord.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/matrix_coord.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/matrix_shape.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/matrix_shape.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/numeric_conversion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/numeric_conversion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/numeric_types.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/numeric_types.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/pipeline/pipeline.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/pipeline/pipeline.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/pitch_linear_coord.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/pitch_linear_coord.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/platform/platform.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/platform/platform.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/predicate_vector.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/predicate_vector.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/quaternion.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/quaternion.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/real.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/real.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/device/reduce_split_k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/device/reduce_split_k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/device/tensor_reduce.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/device/tensor_reduce.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/device/tensor_reduce_affine_contiguous.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/device/tensor_reduce_affine_contiguous.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/device/tensor_reduce_affine_strided.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/device/tensor_reduce_affine_strided.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/kernel/reduce_softmax_final.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/kernel/reduce_softmax_final.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/kernel/reduce_split_k.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/kernel/reduce_split_k.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/kernel/tensor_reduce_affine_contiguous.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/kernel/tensor_reduce_affine_contiguous.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/kernel/tensor_reduce_affine_strided.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/kernel/tensor_reduce_affine_strided.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/thread/reduce.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/thread/reduce.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/thread/reduction_operators.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/thread/reduction_operators.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/reduction/threadblock_swizzle.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/reduction/threadblock_swizzle.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/relatively_equal.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/relatively_equal.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/semaphore.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/semaphore.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/subbyte_reference.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/subbyte_reference.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tensor_coord.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tensor_coord.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tensor_ref.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tensor_ref.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tensor_ref_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tensor_ref_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tensor_view.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tensor_view.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tensor_view_planar_complex.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tensor_view_planar_complex.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/tfloat32.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/tfloat32.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/thread/matrix.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/thread/matrix.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/trace.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/trace.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/collective/sm90_wgmma_transpose.hpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/collective/sm90_wgmma_transpose.hpp -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/pitch_linear_thread_map.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/pitch_linear_thread_map.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/thread/transpose.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/thread/transpose.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/thread/unary_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/thread/unary_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/ell_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/ell_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_scale_bias_vector_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_scale_bias_vector_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_scale_bias_vector_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_scale_bias_vector_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_2dthreadtile.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_2dthreadtile.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_params.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_params.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_triangular_matrix.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator_triangular_matrix.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_2dthreadtile.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_2dthreadtile.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_triangular_matrix.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator_triangular_matrix.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/predicated_vector_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/predicated_vector_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_scale_bias_vector_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_scale_bias_vector_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_pitch_linear.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_pitch_linear.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_pitch_linear_direct_conv.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_pitch_linear_direct_conv.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op_sm80.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op_sm80.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_pitch_linear.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_pitch_linear.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_pitch_linear_2dthreadtile.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_pitch_linear_2dthreadtile.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op_sm70.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op_sm70.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/threadblock/vector_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/threadblock/vector_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/transform/warp/vector_fragment_iterator.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/transform/warp/vector_fragment_iterator.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/uint128.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/uint128.h -------------------------------------------------------------------------------- /third_party/cutlass/include/cutlass/wmma_array.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/third_party/cutlass/include/cutlass/wmma_array.h -------------------------------------------------------------------------------- /tools/performance/performance.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/tools/performance/performance.py -------------------------------------------------------------------------------- /tools/performance/performance.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bruce-Lee-LY/flash_attention_inference/HEAD/tools/performance/performance.sh --------------------------------------------------------------------------------