README.md
pyproject.toml
b12x/__init__.py
b12x.egg-info/PKG-INFO
b12x.egg-info/SOURCES.txt
b12x.egg-info/dependency_links.txt
b12x.egg-info/requires.txt
b12x.egg-info/top_level.txt
b12x/attention/__init__.py
b12x/attention/workspace.py
b12x/attention/_cute/__init__.py
b12x/attention/_cute/copy.py
b12x/attention/_cute/ops.py
b12x/attention/_cute/pipeline.py
b12x/attention/contiguous/README.md
b12x/attention/contiguous/__init__.py
b12x/attention/contiguous/api.py
b12x/attention/contiguous/block_info.py
b12x/attention/contiguous/cute_dsl_utils.py
b12x/attention/contiguous/forward.py
b12x/attention/contiguous/layout_utils.py
b12x/attention/contiguous/mask.py
b12x/attention/contiguous/named_barrier.py
b12x/attention/contiguous/pack_gqa.py
b12x/attention/contiguous/seqlen_info.py
b12x/attention/contiguous/softmax.py
b12x/attention/contiguous/tile_scheduler.py
b12x/attention/indexer/MSA.md
b12x/attention/indexer/__init__.py
b12x/attention/indexer/api.py
b12x/attention/indexer/contiguous_kernel.py
b12x/attention/indexer/fused_indexer.py
b12x/attention/indexer/kernel.py
b12x/attention/indexer/msa_reference.py
b12x/attention/indexer/paged.py
b12x/attention/indexer/persistent_topk.py
b12x/attention/indexer/reference.py
b12x/attention/indexer/schedule_metadata.py
b12x/attention/indexer/scratch.py
b12x/attention/indexer/tiled_topk.py
b12x/attention/mla/__init__.py
b12x/attention/mla/api.py
b12x/attention/mla/compressed_api.py
b12x/attention/mla/compressed_config.py
b12x/attention/mla/compressed_reference.py
b12x/attention/mla/decode_math.py
b12x/attention/mla/io.py
b12x/attention/mla/io_mg.py
b12x/attention/mla/kernel.py
b12x/attention/mla/merge.py
b12x/attention/mla/packed.py
b12x/attention/mla/prefill.py
b12x/attention/mla/prefill_mg.py
b12x/attention/mla/reference.py
b12x/attention/mla/smem.py
b12x/attention/mla/smem_mg.py
b12x/attention/mla/traits.py
b12x/attention/mla/legacy/__init__.py
b12x/attention/mla/legacy/kernel.py
b12x/attention/mla/legacy/kernel_onepass.py
b12x/attention/mla/legacy/split.py
b12x/attention/mla/legacy/traits.py
b12x/attention/paged/__init__.py
b12x/attention/paged/api.py
b12x/attention/paged/forward_extend_generic.py
b12x/attention/paged/forward_paged.py
b12x/attention/paged/graph_replay.py
b12x/attention/paged/merge.py
b12x/attention/paged/planner.py
b12x/attention/paged/reference.py
b12x/attention/paged/traits.py
b12x/attention/paged/workspace.py
b12x/attention/paged/tuning/__init__.py
b12x/attention/paged/tuning/registry.py
b12x/cute/__init__.py
b12x/cute/compiler.py
b12x/cute/fp4.py
b12x/cute/runtime_control.py
b12x/cute/runtime_patches.py
b12x/cute/scratch.py
b12x/cute/utils.py
b12x/distributed/__init__.py
b12x/distributed/_cuda_ipc.py
b12x/distributed/pcie_dcp_a2a.cu
b12x/distributed/pcie_dcp_a2a.py
b12x/distributed/pcie_dma.cu
b12x/distributed/pcie_dma.py
b12x/distributed/pcie_oneshot.cu
b12x/distributed/pcie_oneshot.py
b12x/distributed/pcie_twoshot.cu
b12x/distributed/pcie_twoshot.py
b12x/gemm/__init__.py
b12x/gemm/block_fp8_linear.py
b12x/gemm/dense.py
b12x/gemm/mxfp8_linear.py
b12x/gemm/mxfp8_quant_cute.py
b12x/gemm/wo_projection.py
b12x/gemm/wo_quant_cute.py
b12x/integration/__init__.py
b12x/integration/attention.py
b12x/integration/compressed_scratch.py
b12x/integration/ep_moe.py
b12x/integration/mla.py
b12x/integration/paged_attention_scratch.py
b12x/integration/residual.py
b12x/integration/residual_kernels.py
b12x/integration/scratch.py
b12x/integration/scratch_layout.py
b12x/integration/sparse_mla_scratch.py
b12x/integration/tp_moe.py
b12x/integration/triton_route.py
b12x/moe/__init__.py
b12x/moe/execution.py
b12x/moe/fused/__init__.py
b12x/moe/fused/activations.py
b12x/moe/fused/dynamic.py
b12x/moe/fused/micro.py
b12x/moe/fused/reference.py
b12x/moe/fused/relu2.py
b12x/moe/fused/silu.py
b12x/moe/fused/tiny_decode.py
b12x/moe/fused/w4a8_phase1.py
b12x/moe/fused/w4a8_phase2.py
b12x/moe/fused/w4a16/__init__.py
b12x/moe/fused/w4a16/host.py
b12x/moe/fused/w4a16/kernel.py
b12x/moe/fused/w4a16/prepare.py
b12x/moe/fused/w4a16/route_pack.py
b12x/moe/fused/w4a8/__init__.py
b12x/moe/fused/w4a8/weights.py
b12x/moe/tuning/__init__.py
b12x/moe/tuning/decode.max_active_clusters.py
b12x/moe/tuning/registry.py
b12x/quantization/__init__.py
b12x/quantization/bf16_to_fp4_tma.py
tests/test_attention_contiguous_torch_ref.py
tests/test_attention_cuda_graphs.py
tests/test_attention_mla_api.py
tests/test_attention_mla_compressed.py
tests/test_attention_mla_mxfp8_probe.py
tests/test_attention_mla_reference.py
tests/test_attention_mla_sglang_boundary.py
tests/test_attention_mla_sm120.py
tests/test_attention_msa_e2e.py
tests/test_attention_msa_indexer_api.py
tests/test_attention_msa_indexer_reference.py
tests/test_attention_nsa_backend_sglang_boundary.py
tests/test_attention_nsa_indexer_api.py
tests/test_attention_nsa_indexer_reference.py
tests/test_attention_nsa_indexer_sglang_boundary.py
tests/test_attention_paged_api.py
tests/test_attention_paged_decode_split_graph.py
tests/test_attention_paged_forward.py
tests/test_attention_paged_merge.py
tests/test_attention_paged_msa.py
tests/test_attention_paged_planner.py
tests/test_attention_paged_traits.py
tests/test_b12x_stream_ordering.py
tests/test_b12x_stream_overlap.py
tests/test_benchmark_compressed_mla.py
tests/test_benchmark_mla.py
tests/test_benchmark_moe_flashinfer.py
tests/test_benchmark_paged_attention.py
tests/test_block_fp8_linear_scratch_bindings.py
tests/test_compressed_scratch_bindings.py
tests/test_contiguous_attention_scratch_bindings.py
tests/test_cutlass_runtime_patches.py
tests/test_dense_gemm_expected_m.py
tests/test_ep_moe_api.py
tests/test_fp4_quantization.py
tests/test_fp8_bf16_sanity.py
tests/test_fp8_quant_deepgemm_parity.py
tests/test_fused_indexer.py
tests/test_gemm_block_fp8_linear.py
tests/test_gemm_mxfp8_linear.py
tests/test_gemm_stack.py
tests/test_gemm_wo_projection.py
tests/test_indexer_kernel_bindings.py
tests/test_launch_custom_ops.py
tests/test_mla_kernel_bindings.py
tests/test_moe_eager_prefill.py
tests/test_moe_equivalence.py
tests/test_moe_execution_model.py
tests/test_moe_launch_param_regression.py
tests/test_paged_attention_scratch_bindings.py
tests/test_paged_indexer_integration.py
tests/test_persistent_topk_scratch_bindings.py
tests/test_reference_metrics.py
tests/test_residual_mhc.py
tests/test_residual_scratch_bindings.py
tests/test_runtime_control.py
tests/test_sparse_moe_api.py
tests/test_sparse_moe_qwen_shapes.py
tests/test_tp_moe_prequant_parity.py
tests/test_tp_moe_reference.py
tests/test_tp_moe_relu2_reference.py
tests/test_tp_moe_scratch_bindings.py
tests/test_tp_moe_swigluoai_reference.py
tests/test_tp_moe_w13_layout.py
tests/test_w4a16_e2e.py
tests/test_w4a16_packed_format.py
tests/test_w4a16_reference.py
tests/test_w4a16_route_pack.py
tests/test_w4a8_dynamic_kernel.py
tests/test_w4a8_fragment_probe.py
tests/test_w4a8_mx_tp_moe.py
tests/test_w4a8_primitives.py
tests/test_w4a8_reference.py
tests/test_w4a8_rp_inverse_mapping.py
tests/test_w4a8_tma_probe.py
tests/test_w4a8_tp_moe.py
tests/test_wo_projection_scratch_bindings.py