CHANGELOG.md
CONTRIBUTING.md
LICENSE
MANIFEST.in
NOTICE
README.md
pyproject.toml
setup.py
tqkv/__init__.py
tqkv/_codebooks.py
tqkv/auto_calibrate.py
tqkv/bit_allocator.py
tqkv/calibrate_online.py
tqkv/codec.py
tqkv/core.py
tqkv/calibration/__init__.py
tqkv/calibration/build_distortion_table.py
tqkv/calibration/calibrate_model.py
tqkv/calibration/solve_bits.py
tqkv/csrc/third_party/flash_attention_cute/__init__.py
tqkv/csrc/third_party/flash_attention_cute/ampere_helpers.py
tqkv/csrc/third_party/flash_attention_cute/block_info.py
tqkv/csrc/third_party/flash_attention_cute/block_sparsity.py
tqkv/csrc/third_party/flash_attention_cute/cache_utils.py
tqkv/csrc/third_party/flash_attention_cute/cute_dsl_utils.py
tqkv/csrc/third_party/flash_attention_cute/fa_logging.py
tqkv/csrc/third_party/flash_attention_cute/fast_math.py
tqkv/csrc/third_party/flash_attention_cute/flash_fwd.py
tqkv/csrc/third_party/flash_attention_cute/flash_fwd_sm120.py
tqkv/csrc/third_party/flash_attention_cute/mask.py
tqkv/csrc/third_party/flash_attention_cute/named_barrier.py
tqkv/csrc/third_party/flash_attention_cute/pack_gqa.py
tqkv/csrc/third_party/flash_attention_cute/paged_kv.py
tqkv/csrc/third_party/flash_attention_cute/seqlen_info.py
tqkv/csrc/third_party/flash_attention_cute/softmax.py
tqkv/csrc/third_party/flash_attention_cute/testing.py
tqkv/csrc/third_party/flash_attention_cute/tile_scheduler.py
tqkv/csrc/third_party/flash_attention_cute/utils.py
tqkv/debug/__init__.py
tqkv/debug/oom_snapshot.py
tqkv/integrations/__init__.py
tqkv/integrations/sglang/__init__.py
tqkv/integrations/sglang/_cg.py
tqkv/integrations/sglang/_cold_drain.py
tqkv/integrations/sglang/_metadata.py
tqkv/integrations/sglang/_patch_runner.py
tqkv/integrations/sglang/backend.py
tqkv/integrations/sglang/pool.py
tqkv/integrations/vllm/__init__.py
tqkv/integrations/vllm/_lifecycle.py
tqkv/integrations/vllm/_vendored_fa_metadata_builder.py
tqkv/integrations/vllm/backend.py
tqkv/integrations/vllm/fa4_prefill_adapters.py
tqkv/integrations/vllm/mla_impl.py
tqkv/integrations/vllm/mla_prefill_ref.py
tqkv/integrations/vllm/mla_spec.py
tqkv/integrations/vllm/plugin.py
tqkv/kernels/__init__.py
tqkv/kernels/_cuda_decode_mla.cu
tqkv/kernels/_cuda_decode_mla_splitk.cu
tqkv/kernels/_cuda_decode_splitk.cu
tqkv/kernels/_cuda_decode_splitk_reduce.cuh
tqkv/kernels/_cuda_decode_swa_splitk.cu
tqkv/kernels/_cuda_decode_unified.cu
tqkv/kernels/_cuda_prefill_mla_gather.cu
tqkv/kernels/_decode_common.py
tqkv/kernels/_decode_core.cuh
tqkv/kernels/_jit_common.py
tqkv/kernels/cuda_compress_store.py
tqkv/kernels/cuda_decode.py
tqkv/kernels/cuda_decode_mla.py
tqkv/kernels/cuda_decode_mtp.py
tqkv/kernels/cuda_decode_splitk.py
tqkv/kernels/cuda_decode_swa_splitk.py
tqkv/kernels/cuda_prefill_mla_gather.py
tqkv/kernels/page_metadata.py
tqkv/kernels/precompile.py
tqkv/kernels/triton_kernels.py
tqkv/kernels/cuda/__init__.py
tqkv/kernels/cuda/_ops.py
tqkv/kernels/cuda/tq_kernels.cuh
tqkv/kernels/cuda/tq_ops.cu
tqkv/kernels/cuda/include/tqkv/tq_dequant.cuh
tqkv/kernels/cuda/include/tqkv/tq_inline.cuh
tqkv/kernels/cuda/include/tqkv/tq_norm.cuh
tqkv/kernels/cuda/include/tqkv/tq_page_metadata.cuh
tqkv/kernels/cuda/include/tqkv/tq_paged.cuh
tqkv/kernels/cuda/include/tqkv/tq_unpack.cuh
tqkv/kernels/cuda/include/tqkv/tq_wht.cuh
tqkv/kernels/cuda/include/tqkv/tqkv.cuh
tqkv/kernels/cuda/prefill/__init__.py
tqkv/kernels/cuda/prefill/_tq4_test_codec.py
tqkv/kernels/cuda/prefill/bench_element_size_q_regs.py
tqkv/kernels/cuda/prefill/bench_lpt_sweep.py
tqkv/kernels/cuda/prefill/fa4_tq.py
tqkv/kernels/cuda/prefill/fa4_tq_base.py
tqkv/kernels/cuda/prefill/fa4_tq_dispatch.py
tqkv/kernels/cuda/prefill/fa4_tq_load.py
tqkv/kernels/cuda/prefill/fa4_tq_reference.py
tqkv/kernels/cuda/prefill/fa4_tq_split_d.py
tqkv/kernels/cuda/prefill/triton_prefill_packgqa.py
tqkv/kernels/cuda/prefill/tests/test_fa4_cpasync_spike.py
tqkv/kernels/cuda/prefill/tests/test_fa4_varlen_hellaswag.py
tqkv/kernels/cuda/prefill/tests/test_fa4_varlen_parity.py
tqkv/kernels/cuda/prefill/tests/test_varlen_metadata_cpu.py
tqkv/kernels/experimental/__init__.py
tqkv/kernels/experimental/triton_prefill_paged.py
tqkv/prefill_engines/__init__.py
tqkv/prefill_engines/attend_cuda_v9.py
tqkv/prefill_engines/attend_triton.py
tqkv/prefill_engines/prefill_v9.py
tqkv/prefill_engines/tqkv_prefill_prod.cu
tqkv/runtime/__init__.py
tqkv/runtime/_fa.py
tqkv/runtime/attention.py
tqkv/runtime/autotune.py
tqkv/runtime/cache_views.py
tqkv/runtime/cg_state.py
tqkv/runtime/compress_store.py
tqkv/runtime/core_features.py
tqkv/runtime/decompress.py
tqkv/runtime/layout.py
tqkv/runtime/metadata.py
tqkv/runtime/page_metadata.py
tqkv/runtime/rotation.py
tqkv/runtime/attend/__init__.py
tqkv/runtime/attend/bypass.py
tqkv/runtime/attend/fa4.py
tqkv/runtime/attend/flash_attn.py
tqkv/runtime/attend/mtp.py
tqkv/runtime/attend/native_tq.py
tqkv/runtime/cold/__init__.py
tqkv/runtime/cold/aging_buffer.py
tqkv/runtime/cold/attention.py
tqkv/runtime/cold/block_free_adapter.py
tqkv/runtime/cold/block_free_tracker.py
tqkv/runtime/cold/cg_metadata.py
tqkv/runtime/cold/codec_interface.py
tqkv/runtime/cold/cold_tier.py
tqkv/runtime/cold/compaction.py
tqkv/runtime/cold/config.py
tqkv/runtime/cold/hot_path.py
tqkv/runtime/cold/lse_merge.py
tqkv/runtime/cold/mass_tracker.py
tqkv/runtime/cold/server_integration.py
tqkv/runtime/cold/state.py
tqkv/runtime/cold/tiered_route.py
tqkv/runtime/mla/__init__.py
tqkv/runtime/mla/ops.py
tqkv/runtime/mla/prefill_ref.py
tqkv/runtime/mla/spec.py
turbo_attn.egg-info/PKG-INFO
turbo_attn.egg-info/SOURCES.txt
turbo_attn.egg-info/dependency_links.txt
turbo_attn.egg-info/entry_points.txt
turbo_attn.egg-info/requires.txt
turbo_attn.egg-info/top_level.txt