nvmath-python 1.0.0__cp314-cp314t-win_amd64.whl

nvmath/bindings/cycusparseLt.pxd

@@ -0,0 +1,135 @@
+# Copyright (c) 2024-2026, NVIDIA CORPORATION & AFFILIATES. ALL RIGHTS RESERVED.
+#
+# SPDX-License-Identifier: Apache-2.0
+#
+# This code was automatically generated across versions from 0.7.1 to 0.8.1, generator version 0.3.1.dev1471+g7122059e9. Do not modify it directly.
+# This layer exposes the C header to Cython as-is.
+
+from libc.stdint cimport int64_t, uint8_t, uint32_t, int32_t
+from libc.stdio cimport FILE
+from .cycusparse cimport cusparseStatus_t, _CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR, cusparseOrder_t, cusparseOperation_t
+from ._internal.common_types cimport cudaDataType, cudaDataType_t, cudaStream_t, libraryPropertyType, libraryPropertyType_t, cuComplex, cuDoubleComplex
+
+
+###############################################################################
+# Types (structs, enums, ...)
+###############################################################################
+
+# enums
+ctypedef enum cusparseLtSparsity_t "cusparseLtSparsity_t":
+    CUSPARSELT_SPARSITY_50_PERCENT "CUSPARSELT_SPARSITY_50_PERCENT"
+
+ctypedef enum cusparseLtMatDescAttribute_t "cusparseLtMatDescAttribute_t":
+    CUSPARSELT_MAT_NUM_BATCHES "CUSPARSELT_MAT_NUM_BATCHES"
+    CUSPARSELT_MAT_BATCH_STRIDE "CUSPARSELT_MAT_BATCH_STRIDE"
+
+ctypedef enum cusparseComputeType "cusparseComputeType":
+    CUSPARSE_COMPUTE_32I "CUSPARSE_COMPUTE_32I"
+    CUSPARSE_COMPUTE_16F "CUSPARSE_COMPUTE_16F"
+    CUSPARSE_COMPUTE_32F "CUSPARSE_COMPUTE_32F"
+
+ctypedef enum cusparseLtMatmulDescAttribute_t "cusparseLtMatmulDescAttribute_t":
+    CUSPARSELT_MATMUL_ACTIVATION_RELU "CUSPARSELT_MATMUL_ACTIVATION_RELU"
+    CUSPARSELT_MATMUL_ACTIVATION_RELU_UPPERBOUND "CUSPARSELT_MATMUL_ACTIVATION_RELU_UPPERBOUND"
+    CUSPARSELT_MATMUL_ACTIVATION_RELU_THRESHOLD "CUSPARSELT_MATMUL_ACTIVATION_RELU_THRESHOLD"
+    CUSPARSELT_MATMUL_ACTIVATION_GELU "CUSPARSELT_MATMUL_ACTIVATION_GELU"
+    CUSPARSELT_MATMUL_ACTIVATION_GELU_SCALING "CUSPARSELT_MATMUL_ACTIVATION_GELU_SCALING"
+    CUSPARSELT_MATMUL_ALPHA_VECTOR_SCALING "CUSPARSELT_MATMUL_ALPHA_VECTOR_SCALING"
+    CUSPARSELT_MATMUL_BETA_VECTOR_SCALING "CUSPARSELT_MATMUL_BETA_VECTOR_SCALING"
+    CUSPARSELT_MATMUL_BIAS_STRIDE "CUSPARSELT_MATMUL_BIAS_STRIDE"
+    CUSPARSELT_MATMUL_BIAS_POINTER "CUSPARSELT_MATMUL_BIAS_POINTER"
+    CUSPARSELT_MATMUL_SPARSE_MAT_POINTER "CUSPARSELT_MATMUL_SPARSE_MAT_POINTER"
+    CUSPARSELT_MATMUL_A_SCALE_MODE "CUSPARSELT_MATMUL_A_SCALE_MODE"
+    CUSPARSELT_MATMUL_B_SCALE_MODE "CUSPARSELT_MATMUL_B_SCALE_MODE"
+    CUSPARSELT_MATMUL_C_SCALE_MODE "CUSPARSELT_MATMUL_C_SCALE_MODE"
+    CUSPARSELT_MATMUL_D_SCALE_MODE "CUSPARSELT_MATMUL_D_SCALE_MODE"
+    CUSPARSELT_MATMUL_D_OUT_SCALE_MODE "CUSPARSELT_MATMUL_D_OUT_SCALE_MODE"
+    CUSPARSELT_MATMUL_A_SCALE_POINTER "CUSPARSELT_MATMUL_A_SCALE_POINTER"
+    CUSPARSELT_MATMUL_B_SCALE_POINTER "CUSPARSELT_MATMUL_B_SCALE_POINTER"
+    CUSPARSELT_MATMUL_C_SCALE_POINTER "CUSPARSELT_MATMUL_C_SCALE_POINTER"
+    CUSPARSELT_MATMUL_D_SCALE_POINTER "CUSPARSELT_MATMUL_D_SCALE_POINTER"
+    CUSPARSELT_MATMUL_D_OUT_SCALE_POINTER "CUSPARSELT_MATMUL_D_OUT_SCALE_POINTER"
+
+ctypedef enum cusparseLtMatmulMatrixScale_t "cusparseLtMatmulMatrixScale_t":
+    CUSPARSELT_MATMUL_SCALE_NONE "CUSPARSELT_MATMUL_SCALE_NONE"
+    CUSPARSELT_MATMUL_MATRIX_SCALE_SCALAR_32F "CUSPARSELT_MATMUL_MATRIX_SCALE_SCALAR_32F"
+    CUSPARSELT_MATMUL_MATRIX_SCALE_VEC32_UE4M3 "CUSPARSELT_MATMUL_MATRIX_SCALE_VEC32_UE4M3"
+    CUSPARSELT_MATMUL_MATRIX_SCALE_VEC64_UE8M0 "CUSPARSELT_MATMUL_MATRIX_SCALE_VEC64_UE8M0"
+
+ctypedef enum cusparseLtMatmulAlg_t "cusparseLtMatmulAlg_t":
+    CUSPARSELT_MATMUL_ALG_DEFAULT "CUSPARSELT_MATMUL_ALG_DEFAULT"
+
+ctypedef enum cusparseLtMatmulAlgAttribute_t "cusparseLtMatmulAlgAttribute_t":
+    CUSPARSELT_MATMUL_ALG_CONFIG_ID "CUSPARSELT_MATMUL_ALG_CONFIG_ID"
+    CUSPARSELT_MATMUL_ALG_CONFIG_MAX_ID "CUSPARSELT_MATMUL_ALG_CONFIG_MAX_ID"
+    CUSPARSELT_MATMUL_SEARCH_ITERATIONS "CUSPARSELT_MATMUL_SEARCH_ITERATIONS"
+    CUSPARSELT_MATMUL_SPLIT_K "CUSPARSELT_MATMUL_SPLIT_K"
+    CUSPARSELT_MATMUL_SPLIT_K_MODE "CUSPARSELT_MATMUL_SPLIT_K_MODE"
+    CUSPARSELT_MATMUL_SPLIT_K_BUFFERS "CUSPARSELT_MATMUL_SPLIT_K_BUFFERS"
+
+ctypedef enum cusparseLtSplitKMode_t "cusparseLtSplitKMode_t":
+    CUSPARSELT_INVALID_MODE "CUSPARSELT_INVALID_MODE" = 0
+    CUSPARSELT_SPLIT_K_MODE_ONE_KERNEL "CUSPARSELT_SPLIT_K_MODE_ONE_KERNEL" = 1
+    CUSPARSELT_SPLIT_K_MODE_TWO_KERNELS "CUSPARSELT_SPLIT_K_MODE_TWO_KERNELS" = 2
+    CUSPARSELT_HEURISTIC "CUSPARSELT_HEURISTIC"
+    CUSPARSELT_DATAPARALLEL "CUSPARSELT_DATAPARALLEL"
+    CUSPARSELT_SPLITK "CUSPARSELT_SPLITK"
+    CUSPARSELT_STREAMK "CUSPARSELT_STREAMK"
+
+ctypedef enum cusparseLtPruneAlg_t "cusparseLtPruneAlg_t":
+    CUSPARSELT_PRUNE_SPMMA_TILE "CUSPARSELT_PRUNE_SPMMA_TILE" = 0
+    CUSPARSELT_PRUNE_SPMMA_STRIP "CUSPARSELT_PRUNE_SPMMA_STRIP" = 1
+
+
+# types
+ctypedef struct cusparseLtHandle_t 'cusparseLtHandle_t':
+    uint8_t data[512]
+
+ctypedef struct cusparseLtMatDescriptor_t 'cusparseLtMatDescriptor_t':
+    uint8_t data[512]
+
+ctypedef struct cusparseLtMatmulDescriptor_t 'cusparseLtMatmulDescriptor_t':
+    uint8_t data[512]
+
+ctypedef struct cusparseLtMatmulAlgSelection_t 'cusparseLtMatmulAlgSelection_t':
+    uint8_t data[512]
+
+ctypedef struct cusparseLtMatmulPlan_t 'cusparseLtMatmulPlan_t':
+    uint8_t data[512]
+
+
+###############################################################################
+# Functions
+###############################################################################
+
+cdef const char* cusparseLtGetErrorName(cusparseStatus_t status) except?NULL nogil
+cdef const char* cusparseLtGetErrorString(cusparseStatus_t status) except?NULL nogil
+cdef cusparseStatus_t cusparseLtInit(cusparseLtHandle_t* handle) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtDestroy(const cusparseLtHandle_t* handle) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtGetVersion(const cusparseLtHandle_t* handle, int* version) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtGetProperty(libraryPropertyType propertyType, int* value) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtDenseDescriptorInit(const cusparseLtHandle_t* handle, cusparseLtMatDescriptor_t* matDescr, int64_t rows, int64_t cols, int64_t ld, uint32_t alignment, cudaDataType valueType, cusparseOrder_t order) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtStructuredDescriptorInit(const cusparseLtHandle_t* handle, cusparseLtMatDescriptor_t* matDescr, int64_t rows, int64_t cols, int64_t ld, uint32_t alignment, cudaDataType valueType, cusparseOrder_t order, cusparseLtSparsity_t sparsity) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatDescriptorDestroy(const cusparseLtMatDescriptor_t* matDescr) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatDescSetAttribute(const cusparseLtHandle_t* handle, cusparseLtMatDescriptor_t* matmulDescr, cusparseLtMatDescAttribute_t matAttribute, const void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatDescGetAttribute(const cusparseLtHandle_t* handle, const cusparseLtMatDescriptor_t* matmulDescr, cusparseLtMatDescAttribute_t matAttribute, void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulDescriptorInit(const cusparseLtHandle_t* handle, cusparseLtMatmulDescriptor_t* matmulDescr, cusparseOperation_t opA, cusparseOperation_t opB, const cusparseLtMatDescriptor_t* matA, const cusparseLtMatDescriptor_t* matB, const cusparseLtMatDescriptor_t* matC, const cusparseLtMatDescriptor_t* matD, cusparseComputeType computeType) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulDescSetAttribute(const cusparseLtHandle_t* handle, cusparseLtMatmulDescriptor_t* matmulDescr, cusparseLtMatmulDescAttribute_t matmulAttribute, const void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulDescGetAttribute(const cusparseLtHandle_t* handle, const cusparseLtMatmulDescriptor_t* matmulDescr, cusparseLtMatmulDescAttribute_t matmulAttribute, void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulAlgSelectionInit(const cusparseLtHandle_t* handle, cusparseLtMatmulAlgSelection_t* algSelection, const cusparseLtMatmulDescriptor_t* matmulDescr, cusparseLtMatmulAlg_t alg) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulAlgSetAttribute(const cusparseLtHandle_t* handle, cusparseLtMatmulAlgSelection_t* algSelection, cusparseLtMatmulAlgAttribute_t attribute, const void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulAlgGetAttribute(const cusparseLtHandle_t* handle, const cusparseLtMatmulAlgSelection_t* algSelection, cusparseLtMatmulAlgAttribute_t attribute, void* data, size_t dataSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulGetWorkspace(const cusparseLtHandle_t* handle, const cusparseLtMatmulPlan_t* plan, size_t* workspaceSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulPlanInit(const cusparseLtHandle_t* handle, cusparseLtMatmulPlan_t* plan, const cusparseLtMatmulDescriptor_t* matmulDescr, const cusparseLtMatmulAlgSelection_t* algSelection) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulPlanDestroy(const cusparseLtMatmulPlan_t* plan) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmul(const cusparseLtHandle_t* handle, const cusparseLtMatmulPlan_t* plan, const void* alpha, const void* d_A, const void* d_B, const void* beta, const void* d_C, void* d_D, void* workspace, cudaStream_t* streams, int32_t numStreams) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulSearch(const cusparseLtHandle_t* handle, cusparseLtMatmulPlan_t* plan, const void* alpha, const void* d_A, const void* d_B, const void* beta, const void* d_C, void* d_D, void* workspace, cudaStream_t* streams, int32_t numStreams) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMAPrune(const cusparseLtHandle_t* handle, const cusparseLtMatmulDescriptor_t* matmulDescr, const void* d_in, void* d_out, cusparseLtPruneAlg_t pruneAlg, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMAPruneCheck(const cusparseLtHandle_t* handle, const cusparseLtMatmulDescriptor_t* matmulDescr, const void* d_in, int* valid, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMAPrune2(const cusparseLtHandle_t* handle, const cusparseLtMatDescriptor_t* sparseMatDescr, int isSparseA, cusparseOperation_t op, const void* d_in, void* d_out, cusparseLtPruneAlg_t pruneAlg, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMAPruneCheck2(const cusparseLtHandle_t* handle, const cusparseLtMatDescriptor_t* sparseMatDescr, int isSparseA, cusparseOperation_t op, const void* d_in, int* d_valid, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMACompressedSize(const cusparseLtHandle_t* handle, const cusparseLtMatmulPlan_t* plan, size_t* compressedSize, size_t* compressedBufferSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMACompress(const cusparseLtHandle_t* handle, const cusparseLtMatmulPlan_t* plan, const void* d_dense, void* d_compressed, void* d_compressed_buffer, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMACompressedSize2(const cusparseLtHandle_t* handle, const cusparseLtMatDescriptor_t* sparseMatDescr, size_t* compressedSize, size_t* compressedBufferSize) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtSpMMACompress2(const cusparseLtHandle_t* handle, const cusparseLtMatDescriptor_t* sparseMatDescr, int isSparseA, cusparseOperation_t op, const void* d_dense, void* d_compressed, void* d_compressed_buffer, cudaStream_t stream) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cusparseStatus_t cusparseLtMatmulAlgSelectionDestroy(const cusparseLtMatmulAlgSelection_t* algSelection) except?_CUSPARSESTATUS_T_INTERNAL_LOADING_ERROR nogil

nvmath/bindings/cycutensor.pxd

@@ -0,0 +1,189 @@
+# Copyright (c) 2025-2026, NVIDIA CORPORATION & AFFILIATES. ALL RIGHTS RESERVED.
+#
+# SPDX-License-Identifier: Apache-2.0
+#
+# This code was automatically generated across versions from 2.5.0 to 2.6.0, generator version 0.3.1.dev1471+g7122059e9. Do not modify it directly.
+# This layer exposes the C header to Cython as-is.
+
+from libc.stdint cimport int32_t, int64_t, uint32_t, uint64_t
+from libc.stdio cimport FILE
+from ._internal.common_types cimport cudaDataType, cudaDataType_t, cudaStream_t, libraryPropertyType, libraryPropertyType_t
+
+
+###############################################################################
+# Types (structs, enums, ...)
+###############################################################################
+
+# enums
+ctypedef enum cutensorOperator_t "cutensorOperator_t":
+    CUTENSOR_OP_IDENTITY "CUTENSOR_OP_IDENTITY" = 1
+    CUTENSOR_OP_SQRT "CUTENSOR_OP_SQRT" = 2
+    CUTENSOR_OP_RELU "CUTENSOR_OP_RELU" = 8
+    CUTENSOR_OP_CONJ "CUTENSOR_OP_CONJ" = 9
+    CUTENSOR_OP_RCP "CUTENSOR_OP_RCP" = 10
+    CUTENSOR_OP_SIGMOID "CUTENSOR_OP_SIGMOID" = 11
+    CUTENSOR_OP_TANH "CUTENSOR_OP_TANH" = 12
+    CUTENSOR_OP_EXP "CUTENSOR_OP_EXP" = 22
+    CUTENSOR_OP_LOG "CUTENSOR_OP_LOG" = 23
+    CUTENSOR_OP_ABS "CUTENSOR_OP_ABS" = 24
+    CUTENSOR_OP_NEG "CUTENSOR_OP_NEG" = 25
+    CUTENSOR_OP_SIN "CUTENSOR_OP_SIN" = 26
+    CUTENSOR_OP_COS "CUTENSOR_OP_COS" = 27
+    CUTENSOR_OP_TAN "CUTENSOR_OP_TAN" = 28
+    CUTENSOR_OP_SINH "CUTENSOR_OP_SINH" = 29
+    CUTENSOR_OP_COSH "CUTENSOR_OP_COSH" = 30
+    CUTENSOR_OP_ASIN "CUTENSOR_OP_ASIN" = 31
+    CUTENSOR_OP_ACOS "CUTENSOR_OP_ACOS" = 32
+    CUTENSOR_OP_ATAN "CUTENSOR_OP_ATAN" = 33
+    CUTENSOR_OP_ASINH "CUTENSOR_OP_ASINH" = 34
+    CUTENSOR_OP_ACOSH "CUTENSOR_OP_ACOSH" = 35
+    CUTENSOR_OP_ATANH "CUTENSOR_OP_ATANH" = 36
+    CUTENSOR_OP_CEIL "CUTENSOR_OP_CEIL" = 37
+    CUTENSOR_OP_FLOOR "CUTENSOR_OP_FLOOR" = 38
+    CUTENSOR_OP_MISH "CUTENSOR_OP_MISH" = 39
+    CUTENSOR_OP_SWISH "CUTENSOR_OP_SWISH" = 40
+    CUTENSOR_OP_SOFT_PLUS "CUTENSOR_OP_SOFT_PLUS" = 41
+    CUTENSOR_OP_SOFT_SIGN "CUTENSOR_OP_SOFT_SIGN" = 42
+    CUTENSOR_OP_ADD "CUTENSOR_OP_ADD" = 3
+    CUTENSOR_OP_MUL "CUTENSOR_OP_MUL" = 5
+    CUTENSOR_OP_MAX "CUTENSOR_OP_MAX" = 6
+    CUTENSOR_OP_MIN "CUTENSOR_OP_MIN" = 7
+    CUTENSOR_OP_UNKNOWN "CUTENSOR_OP_UNKNOWN" = 126
+
+ctypedef enum cutensorStatus_t "cutensorStatus_t":
+    CUTENSOR_STATUS_SUCCESS "CUTENSOR_STATUS_SUCCESS" = 0
+    CUTENSOR_STATUS_NOT_INITIALIZED "CUTENSOR_STATUS_NOT_INITIALIZED" = 1
+    CUTENSOR_STATUS_ALLOC_FAILED "CUTENSOR_STATUS_ALLOC_FAILED" = 3
+    CUTENSOR_STATUS_INVALID_VALUE "CUTENSOR_STATUS_INVALID_VALUE" = 7
+    CUTENSOR_STATUS_ARCH_MISMATCH "CUTENSOR_STATUS_ARCH_MISMATCH" = 8
+    CUTENSOR_STATUS_MAPPING_ERROR "CUTENSOR_STATUS_MAPPING_ERROR" = 11
+    CUTENSOR_STATUS_EXECUTION_FAILED "CUTENSOR_STATUS_EXECUTION_FAILED" = 13
+    CUTENSOR_STATUS_INTERNAL_ERROR "CUTENSOR_STATUS_INTERNAL_ERROR" = 14
+    CUTENSOR_STATUS_NOT_SUPPORTED "CUTENSOR_STATUS_NOT_SUPPORTED" = 15
+    CUTENSOR_STATUS_LICENSE_ERROR "CUTENSOR_STATUS_LICENSE_ERROR" = 16
+    CUTENSOR_STATUS_CUBLAS_ERROR "CUTENSOR_STATUS_CUBLAS_ERROR" = 17
+    CUTENSOR_STATUS_CUDA_ERROR "CUTENSOR_STATUS_CUDA_ERROR" = 18
+    CUTENSOR_STATUS_INSUFFICIENT_WORKSPACE "CUTENSOR_STATUS_INSUFFICIENT_WORKSPACE" = 19
+    CUTENSOR_STATUS_INSUFFICIENT_DRIVER "CUTENSOR_STATUS_INSUFFICIENT_DRIVER" = 20
+    CUTENSOR_STATUS_IO_ERROR "CUTENSOR_STATUS_IO_ERROR" = 21
+    _CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR "_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR" = -42
+
+ctypedef enum cutensorAlgo_t "cutensorAlgo_t":
+    CUTENSOR_ALGO_DEFAULT_PATIENT "CUTENSOR_ALGO_DEFAULT_PATIENT" = -(6)
+    CUTENSOR_ALGO_GETT "CUTENSOR_ALGO_GETT" = -(4)
+    CUTENSOR_ALGO_TGETT "CUTENSOR_ALGO_TGETT" = -(3)
+    CUTENSOR_ALGO_TTGT "CUTENSOR_ALGO_TTGT" = -(2)
+    CUTENSOR_ALGO_DEFAULT "CUTENSOR_ALGO_DEFAULT" = -(1)
+
+ctypedef enum cutensorWorksizePreference_t "cutensorWorksizePreference_t":
+    CUTENSOR_WORKSPACE_MIN "CUTENSOR_WORKSPACE_MIN" = 1
+    CUTENSOR_WORKSPACE_DEFAULT "CUTENSOR_WORKSPACE_DEFAULT" = 2
+    CUTENSOR_WORKSPACE_MAX "CUTENSOR_WORKSPACE_MAX" = 3
+
+ctypedef enum cutensorOperationDescriptorAttribute_t "cutensorOperationDescriptorAttribute_t":
+    CUTENSOR_OPERATION_DESCRIPTOR_TAG "CUTENSOR_OPERATION_DESCRIPTOR_TAG" = 0
+    CUTENSOR_OPERATION_DESCRIPTOR_SCALAR_TYPE "CUTENSOR_OPERATION_DESCRIPTOR_SCALAR_TYPE" = 1
+    CUTENSOR_OPERATION_DESCRIPTOR_FLOPS "CUTENSOR_OPERATION_DESCRIPTOR_FLOPS" = 2
+    CUTENSOR_OPERATION_DESCRIPTOR_MOVED_BYTES "CUTENSOR_OPERATION_DESCRIPTOR_MOVED_BYTES" = 3
+    CUTENSOR_OPERATION_DESCRIPTOR_PADDING_LEFT "CUTENSOR_OPERATION_DESCRIPTOR_PADDING_LEFT" = 4
+    CUTENSOR_OPERATION_DESCRIPTOR_PADDING_RIGHT "CUTENSOR_OPERATION_DESCRIPTOR_PADDING_RIGHT" = 5
+    CUTENSOR_OPERATION_DESCRIPTOR_PADDING_VALUE "CUTENSOR_OPERATION_DESCRIPTOR_PADDING_VALUE" = 6
+
+ctypedef enum cutensorPlanPreferenceAttribute_t "cutensorPlanPreferenceAttribute_t":
+    CUTENSOR_PLAN_PREFERENCE_AUTOTUNE_MODE "CUTENSOR_PLAN_PREFERENCE_AUTOTUNE_MODE" = 0
+    CUTENSOR_PLAN_PREFERENCE_CACHE_MODE "CUTENSOR_PLAN_PREFERENCE_CACHE_MODE" = 1
+    CUTENSOR_PLAN_PREFERENCE_INCREMENTAL_COUNT "CUTENSOR_PLAN_PREFERENCE_INCREMENTAL_COUNT" = 2
+    CUTENSOR_PLAN_PREFERENCE_ALGO "CUTENSOR_PLAN_PREFERENCE_ALGO" = 3
+    CUTENSOR_PLAN_PREFERENCE_KERNEL_RANK "CUTENSOR_PLAN_PREFERENCE_KERNEL_RANK" = 4
+    CUTENSOR_PLAN_PREFERENCE_JIT "CUTENSOR_PLAN_PREFERENCE_JIT" = 5
+    CUTENSOR_PLAN_PREFERENCE_GPU_ARCH "CUTENSOR_PLAN_PREFERENCE_GPU_ARCH" = 6
+
+ctypedef enum cutensorAutotuneMode_t "cutensorAutotuneMode_t":
+    CUTENSOR_AUTOTUNE_MODE_NONE "CUTENSOR_AUTOTUNE_MODE_NONE" = 0
+    CUTENSOR_AUTOTUNE_MODE_INCREMENTAL "CUTENSOR_AUTOTUNE_MODE_INCREMENTAL" = 1
+
+ctypedef enum cutensorJitMode_t "cutensorJitMode_t":
+    CUTENSOR_JIT_MODE_NONE "CUTENSOR_JIT_MODE_NONE" = 0
+    CUTENSOR_JIT_MODE_DEFAULT "CUTENSOR_JIT_MODE_DEFAULT" = 1
+
+ctypedef enum cutensorCacheMode_t "cutensorCacheMode_t":
+    CUTENSOR_CACHE_MODE_NONE "CUTENSOR_CACHE_MODE_NONE" = 0
+    CUTENSOR_CACHE_MODE_PEDANTIC "CUTENSOR_CACHE_MODE_PEDANTIC" = 1
+
+ctypedef enum cutensorPlanAttribute_t "cutensorPlanAttribute_t":
+    CUTENSOR_PLAN_REQUIRED_WORKSPACE "CUTENSOR_PLAN_REQUIRED_WORKSPACE" = 0
+
+
+# types
+ctypedef cudaDataType_t cutensorDataType_t 'cutensorDataType_t'
+
+ctypedef void* cutensorComputeDescriptor_t 'cutensorComputeDescriptor_t'
+
+ctypedef void* cutensorOperationDescriptor_t 'cutensorOperationDescriptor_t'
+
+ctypedef void* cutensorPlan_t 'cutensorPlan_t'
+
+ctypedef void* cutensorPlanPreference_t 'cutensorPlanPreference_t'
+
+ctypedef void* cutensorHandle_t 'cutensorHandle_t'
+
+ctypedef void* cutensorTensorDescriptor_t 'cutensorTensorDescriptor_t'
+
+ctypedef void* cutensorBlockSparseTensorDescriptor_t 'cutensorBlockSparseTensorDescriptor_t'
+
+ctypedef void (*cutensorLoggerCallback_t 'cutensorLoggerCallback_t')(
+    int32_t logLevel,
+    const char* functionName,
+    const char* message
+)
+
+
+###############################################################################
+# Functions
+###############################################################################
+
+cdef cutensorStatus_t cutensorCreate(cutensorHandle_t* handle) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroy(cutensorHandle_t handle) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorHandleResizePlanCache(cutensorHandle_t handle, const uint32_t numEntries) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorHandleWritePlanCacheToFile(const cutensorHandle_t handle, const char filename[]) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorHandleReadPlanCacheFromFile(cutensorHandle_t handle, const char filename[], uint32_t* numCachelinesRead) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorWriteKernelCacheToFile(const cutensorHandle_t handle, const char filename[]) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorReadKernelCacheFromFile(cutensorHandle_t handle, const char filename[]) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateTensorDescriptor(const cutensorHandle_t handle, cutensorTensorDescriptor_t* desc, const uint32_t numModes, const int64_t extent[], const int64_t stride[], cudaDataType_t dataType, uint32_t alignmentRequirement) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroyTensorDescriptor(cutensorTensorDescriptor_t desc) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateElementwiseTrinary(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descB, const int32_t modeB[], cutensorOperator_t opB, const cutensorTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorTensorDescriptor_t descD, const int32_t modeD[], cutensorOperator_t opAB, cutensorOperator_t opABC, const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorElementwiseTrinaryExecute(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, const void* beta, const void* B, const void* gamma, const void* C, void* D, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateElementwiseBinary(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorTensorDescriptor_t descD, const int32_t modeD[], cutensorOperator_t opAC, const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorElementwiseBinaryExecute(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, const void* gamma, const void* C, void* D, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreatePermutation(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descB, const int32_t modeB[], const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorPermute(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, void* B, const cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateContraction(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descB, const int32_t modeB[], cutensorOperator_t opB, const cutensorTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorTensorDescriptor_t descD, const int32_t modeD[], const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroyOperationDescriptor(cutensorOperationDescriptor_t desc) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorOperationDescriptorSetAttribute(const cutensorHandle_t handle, cutensorOperationDescriptor_t desc, cutensorOperationDescriptorAttribute_t attr, const void* buf, size_t sizeInBytes) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorOperationDescriptorGetAttribute(const cutensorHandle_t handle, cutensorOperationDescriptor_t desc, cutensorOperationDescriptorAttribute_t attr, void* buf, size_t sizeInBytes) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreatePlanPreference(const cutensorHandle_t handle, cutensorPlanPreference_t* pref, cutensorAlgo_t algo, cutensorJitMode_t jitMode) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroyPlanPreference(cutensorPlanPreference_t pref) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorPlanPreferenceGetAttribute(const cutensorHandle_t handle, cutensorPlanPreference_t pref, cutensorPlanPreferenceAttribute_t attr, void* buf, size_t sizeInBytes) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorPlanPreferenceSetAttribute(const cutensorHandle_t handle, cutensorPlanPreference_t pref, cutensorPlanPreferenceAttribute_t attr, const void* buf, size_t sizeInBytes) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorPlanGetAttribute(const cutensorHandle_t handle, const cutensorPlan_t plan, cutensorPlanAttribute_t attr, void* buf, size_t sizeInBytes) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorEstimateWorkspaceSize(const cutensorHandle_t handle, const cutensorOperationDescriptor_t desc, const cutensorPlanPreference_t planPref, const cutensorWorksizePreference_t workspacePref, uint64_t* workspaceSizeEstimate) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreatePlan(const cutensorHandle_t handle, cutensorPlan_t* plan, const cutensorOperationDescriptor_t desc, const cutensorPlanPreference_t pref, uint64_t workspaceSizeLimit) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroyPlan(cutensorPlan_t plan) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorContract(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, const void* B, const void* beta, const void* C, void* D, void* workspace, uint64_t workspaceSize, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateReduction(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorTensorDescriptor_t descD, const int32_t modeD[], cutensorOperator_t opReduce, const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorReduce(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, const void* beta, const void* C, void* D, void* workspace, uint64_t workspaceSize, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateContractionTrinary(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorTensorDescriptor_t descB, const int32_t modeB[], cutensorOperator_t opB, const cutensorTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorTensorDescriptor_t descD, const int32_t modeD[], cutensorOperator_t opD, const cutensorTensorDescriptor_t descE, const int32_t modeE[], const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorContractTrinary(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* A, const void* B, const void* C, const void* beta, const void* D, void* E, void* workspace, uint64_t workspaceSize, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateBlockSparseTensorDescriptor(cutensorHandle_t handle, cutensorBlockSparseTensorDescriptor_t* desc, const uint32_t numModes, const uint64_t numNonZeroBlocks, const uint32_t numSectionsPerMode[], const int64_t extent[], const int32_t nonZeroCoordinates[], const int64_t stride[], cudaDataType_t dataType) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorDestroyBlockSparseTensorDescriptor(cutensorBlockSparseTensorDescriptor_t desc) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorCreateBlockSparseContraction(const cutensorHandle_t handle, cutensorOperationDescriptor_t* desc, const cutensorBlockSparseTensorDescriptor_t descA, const int32_t modeA[], cutensorOperator_t opA, const cutensorBlockSparseTensorDescriptor_t descB, const int32_t modeB[], cutensorOperator_t opB, const cutensorBlockSparseTensorDescriptor_t descC, const int32_t modeC[], cutensorOperator_t opC, const cutensorBlockSparseTensorDescriptor_t descD, const int32_t modeD[], const cutensorComputeDescriptor_t descCompute) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorBlockSparseContract(const cutensorHandle_t handle, const cutensorPlan_t plan, const void* alpha, const void* const A[], const void* const B[], const void* beta, const void* const C[], void* const D[], void* workspace, uint64_t workspaceSize, cudaStream_t stream) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef const char* cutensorGetErrorString(const cutensorStatus_t error) except?NULL nogil
+cdef size_t cutensorGetVersion() except?0 nogil
+cdef size_t cutensorGetCudartVersion() except?0 nogil
+cdef cutensorStatus_t cutensorLoggerSetCallback(cutensorLoggerCallback_t callback) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorLoggerSetFile(FILE* file) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorLoggerOpenFile(const char* logFile) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorLoggerSetLevel(int32_t level) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorLoggerSetMask(int32_t mask) except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil
+cdef cutensorStatus_t cutensorLoggerForceDisable() except?_CUTENSORSTATUS_T_INTERNAL_LOADING_ERROR nogil