DOCUMENTATION.md
LICENSE
MANIFEST.in
README.md
_C.cp312-win_amd64.pyd
__init__.py
__main__.py
bench_attn.py
bench_bwd.py
bench_fp8_gemm_stress.py
build.ps1
build.sh
cdna_compat.py
demo.py
device_info.py
diagnose_fp8_crash.py
e2e_iq2s_iq1m.py
hip_iquant_util.h
hip_quant_types.h
hip_quant_util.h
hip_quantize.cpp
hip_quantize.dll
hip_quantize_rocm721.dll
pyproject.toml
setup.py
setup_torch.py
test_backward_correctness.py
test_backward_math.py
test_fp8_gemm.py
test_fp8_gemm_debug.py
test_wmma_layout.py
torch_api.py
./_C.cp312-win_amd64.pyd
./__init__.py
./__main__.py
./bench_attn.py
./bench_bwd.py
./bench_fp8_gemm_stress.py
./build.ps1
./build.sh
./cdna_compat.py
./demo.py
./device_info.py
./diagnose_fp8_crash.py
./e2e_iq2s_iq1m.py
./hip_iquant_util.h
./hip_quant_types.h
./hip_quant_util.h
./hip_quantize.cpp
./hip_quantize.dll
./hip_quantize_rocm721.dll
./setup.py
./setup_torch.py
./test_backward_correctness.py
./test_backward_math.py
./test_fp8_gemm.py
./test_fp8_gemm_debug.py
./test_wmma_layout.py
./torch_api.py
./codebooks/iq1_s.bin
./codebooks/iq2_s.bin
./codebooks/iq2_xs.bin
./codebooks/iq2_xxs.bin
./codebooks/iq3_s.bin
./codebooks/iq3_xxs.bin
./kernels/dequant_to_fp8.cu
./kernels/direct_fp8_quant.cu
./kernels/fp8_expand.cu
./kernels/fp8_gemm_test.cu
./kernels/quant_aq2.cu
./kernels/quant_bf16.cu
./kernels/quant_f8_e4m3.cu
./kernels/quant_f8_e5m2.cu
./kernels/quant_hq2.cu
./kernels/quant_iq1_m.cu
./kernels/quant_iq1_s.cu
./kernels/quant_iq2_s.cu
./kernels/quant_iq2_xs.cu
./kernels/quant_iq2_xxs.cu
./kernels/quant_iq3_s.cu
./kernels/quant_iq3_xxs.cu
./kernels/quant_iq4_nl.cu
./kernels/quant_iq4_xs.cu
./kernels/quant_q1_0.cu
./kernels/quant_q2_0.cu
./kernels/quant_q2_K.cu
./kernels/quant_q3_K.cu
./kernels/quant_q4_0.cu
./kernels/quant_q4_1.cu
./kernels/quant_q4_K.cu
./kernels/quant_q5_0.cu
./kernels/quant_q5_1.cu
./kernels/quant_q5_K.cu
./kernels/quant_q6_K.cu
./kernels/quant_q8_0.cu
./kernels/quant_q8_1.cu
./kernels/quant_q8_K.cu
./kernels/quant_tq1_0.cu
./kernels/quant_tq2_0.cu
./kernels/wmma_diag.cu
./tests/bench_attn_compare.py
./tests/test_aq2.py
./tests/test_bf16_numpy.py
./tests/test_compat.py
./tests/test_dequant_fp8.py
./tests/test_direct_fp8_quant.py
./tests/test_hq2.py
./tests/test_hq2_portable.py
./tests/test_hq8_g128.py
./tests/test_iq2s_iq1m_byte_exact.py
./tests/test_iquant_parallel.py
./tests/test_lazy_torch_import.py
./tests/test_mixed_policy.py
./tests/test_mixed_sensitivity.py
./tests/test_mxfp4_compat.py
./tests/test_mxfp4_native_contract.py
./tests/test_patch_hq_mixed_archive.py
./tests/test_pipeline.py
./tests/test_q_to_fp8_gpu.py
./tests/test_qol.py
./tests/test_safetensors_numpy.py
./tests/test_v8.py
./tests/test_wmma_diag.py
./tests/test_wmma_layout.py
./tests/torch/bench_compare.py
./tests/torch/bench_fp8.py
./tests/torch/conftest.py
./tests/torch/test_fp8.py
./tests/torch/test_fp8_v2.py
./tests/torch/test_hip_graph.py
./tests/torch/test_mxfp4_native_contract_torch.py
./tests/torch/test_mxfp4_to_fp8_torch.py
./tests/torch/test_q_to_fp8_torch.py
./tests/torch/test_quant_transpose.py
./tests/torch/test_wave_attn_backward_math.py
./torch_ext/current_stream.hip
./torch_ext/fp8_linear_kernels.hip
./torch_ext/fp8_linear_kernels_v2.hip
./torch_ext/fp8_quant_kernels.hip
./torch_ext/hq2_linear_kernels.hip
./torch_ext/hq3_linear_kernels.hip
./torch_ext/hq8_g128_linear_kernels.hip
./torch_ext/mxfp4_to_fp8_kernels.hip
./torch_ext/pytorch_bindings.cpp
./torch_ext/pytorch_bindings_hip.cpp
./torch_ext/q_to_fp8_kernels.hip
./torch_ext/wave_attn.hip
./torch_ext/wave_attn_backward.hip
./torch_ext/wave_attn_diag.hip
codebooks/iq1_s.bin
codebooks/iq2_s.bin
codebooks/iq2_xs.bin
codebooks/iq2_xxs.bin
codebooks/iq3_s.bin
codebooks/iq3_xxs.bin
hip_quant.egg-info/PKG-INFO
hip_quant.egg-info/SOURCES.txt
hip_quant.egg-info/dependency_links.txt
hip_quant.egg-info/entry_points.txt
hip_quant.egg-info/not-zip-safe
hip_quant.egg-info/requires.txt
hip_quant.egg-info/top_level.txt
hq2/__init__.py
hq2/__main__.py
hq2/analyzer.py
hq2/api.py
hq2/archive.py
hq2/backends.py
hq2/format.py
hq2/hq3.py
hq2/hq8.py
hq2/mixed_policy.py
hq2/raw.py
hq2/safetensors_numpy.py
hq2/sensitivity.py
hq2/torch_inference.py
hq2/transformers_loader.py
kernels/dequant_to_fp8.cu
kernels/direct_fp8_quant.cu
kernels/fp8_expand.cu
kernels/fp8_gemm_test.cu
kernels/quant_aq2.cu
kernels/quant_bf16.cu
kernels/quant_f8_e4m3.cu
kernels/quant_f8_e5m2.cu
kernels/quant_hq2.cu
kernels/quant_iq1_m.cu
kernels/quant_iq1_s.cu
kernels/quant_iq2_s.cu
kernels/quant_iq2_xs.cu
kernels/quant_iq2_xxs.cu
kernels/quant_iq3_s.cu
kernels/quant_iq3_xxs.cu
kernels/quant_iq4_nl.cu
kernels/quant_iq4_xs.cu
kernels/quant_q1_0.cu
kernels/quant_q2_0.cu
kernels/quant_q2_K.cu
kernels/quant_q3_K.cu
kernels/quant_q4_0.cu
kernels/quant_q4_1.cu
kernels/quant_q4_K.cu
kernels/quant_q5_0.cu
kernels/quant_q5_1.cu
kernels/quant_q5_K.cu
kernels/quant_q6_K.cu
kernels/quant_q8_0.cu
kernels/quant_q8_1.cu
kernels/quant_q8_K.cu
kernels/quant_tq1_0.cu
kernels/quant_tq2_0.cu
kernels/wmma_diag.cu
tests/bench_attn_compare.py
tests/test_aq2.py
tests/test_bf16_numpy.py
tests/test_compat.py
tests/test_dequant_fp8.py
tests/test_direct_fp8_quant.py
tests/test_hq2.py
tests/test_hq2_portable.py
tests/test_hq8_g128.py
tests/test_iq2s_iq1m_byte_exact.py
tests/test_iquant_parallel.py
tests/test_lazy_torch_import.py
tests/test_mixed_policy.py
tests/test_mixed_sensitivity.py
tests/test_mxfp4_compat.py
tests/test_mxfp4_native_contract.py
tests/test_patch_hq_mixed_archive.py
tests/test_pipeline.py
tests/test_q_to_fp8_gpu.py
tests/test_qol.py
tests/test_safetensors_numpy.py
tests/test_v8.py
tests/test_wmma_diag.py
tests/test_wmma_layout.py
tests/torch/bench_compare.py
tests/torch/bench_fp8.py
tests/torch/conftest.py
tests/torch/test_fp8.py
tests/torch/test_fp8_v2.py
tests/torch/test_hip_graph.py
tests/torch/test_mxfp4_native_contract_torch.py
tests/torch/test_mxfp4_to_fp8_torch.py
tests/torch/test_q_to_fp8_torch.py
tests/torch/test_quant_transpose.py
tests/torch/test_wave_attn_backward_math.py
torch_ext/current_stream.hip
torch_ext/fp8_linear_kernels.hip
torch_ext/fp8_linear_kernels_v2.hip
torch_ext/fp8_quant_kernels.hip
torch_ext/hq2_linear_kernels.hip
torch_ext/hq3_linear_kernels.hip
torch_ext/hq8_g128_linear_kernels.hip
torch_ext/mxfp4_to_fp8_kernels.hip
torch_ext/pytorch_bindings.cpp
torch_ext/pytorch_bindings_hip.cpp
torch_ext/q_to_fp8_kernels.hip
torch_ext/wave_attn.hip
torch_ext/wave_attn_backward.hip
torch_ext/wave_attn_diag.hip