Sj NddlZddlmZddlmZddlmZddlZddlm Z m Z ddl m Z m Z mZmZmZmZmZmZmZmZmZmZmZgdZedd ZGd d ej6Zd ej:fd ej6dededefdZ GddeZ!GddeZ"d a#dZ$d a%dZ&y)N) namedtuple)Callable)Any))sparse_semi_structured_from_dense_cutlass'sparse_semi_structured_to_dense_cutlass) fallback_dispatchersemi_sparse_addmmsemi_sparse_clonesemi_sparse_detachsemi_sparse_indicessemi_sparse_linearsemi_sparse_mmsemi_sparse_scaled_mm semi_sparse_tsemi_sparse_tosemi_sparse_to_copysemi_sparse_valuessemi_sparse_view)SparseSemiStructuredTensor!SparseSemiStructuredTensorCUTLASS$SparseSemiStructuredTensorCUSPARSELTto_sparse_semi_structured_SEMI_STRUCTURED_SPARSE_CONFIGz=sparse_min_rows sparse_min_cols dense_min_rows dense_min_colsceZdZUdZdZeed<eeje fed<dZ e ed<dZ e ed<dZe ed<eed <eeefed <ej"d zed <ej"d zed <ej"d zed<ej"d zed<ej"d zed<e ed<eed<gdZe d'dej(d ej"d zd ej"d zdej"d zdej"d zdej"d zde dede fdZdefdZdeeeeej(e ee fffdZedeej(e ee fdej"fdZej8j:ZedefdZ ed(d)dZ!edej"dd fdZ"d Z#eefdej"d!eddfd"Z$d d#d$ej"d%ej"d zdej"fd&Z%y )*ra This class implements semi-structured sparsity as a Tensor subclass. Semi-structured sparsity describes a sparsity pattern where n in every 2n elements are sparse, depending on the datatype. It is also referred to as 2:4 sparsity or fine-grained structured sparsity. There are two backends available for semi_structred sparsity, either cuSPARSELt or CUTLASS. This class is meant to serve as a base class for both implementations. SparseSemiStructuredCUTLASS and SparseSemiStructuredCUSPARSELT both inherit from this class and define three backend-specific items. Note that as such, this class cannot be instantiated directly. -`_DTYPE_SHAPE_CONSTRAINTS` - A dictionary holding backend specific dense/sparse min shape constraints - `def from_dense()` - backend specific compression routines - `def _mm()` - backend specific mm op (either torch._cslt_sparse_mm or torch._sparse_semi_structured_(mm|addmm)) r_DEFAULT_ALG_ID_DTYPE_SHAPE_CONSTRAINTSF_FORCE_CUTLASS_FUSE_TRANSPOSE_PROTOTYPE_WARNING_SHOWNBACKENDSPARSE_DISPATCHNpackedmetapacked_tmeta_tcompressed_swizzled_bitmaskfuse_transpose_cusparseltalg_id_cusparselt)r"r#r$r%r&shape requires_gradc |jsRtjdtdd|_|j t j j|||} n||} n tdt jj||| j| j| j| } || _|| _|| _|| _|| _|| _|| _| S)a0 Create a new instance of the tensor subclass from the compressed sparse representation. We have the option to create the subclass with the compressed representations of both X and X', for training. For inference, we only need a single representation (either X or X'), while the corresponding other set will be None. Depending on the backend selected, certain fields will be set to None. (CUSPARSELT vs CUTLASS) Args: shape: The shape of the original dense tensor packed: The compressed representation of the original dense tensor meta: The metadata of the original dense tensor, if it is stored separately packed_t: The compressed representation of the transposed original dense tensor meta_t: The metadata of the transposed original dense tensor, if it is stored separately compressed_swizzled_bitmask: The masks used by the CUTLASS backend to determine which threads should participate in the computation. Used for pointwise ops. fuse_transpose_cusparselt: When running with cuSPARSELt, we have the option to fuse a transposition with a matmul, which is useful in the case of 2:4 sparse training. alg_id_cusparselt: The algorithm id to use when using cuSPARSELT, will have effect on performance Returns: torch.Tensor: A torch.Tensor wrapper subclass. Raises: ValueError: If all of the tensor arguments are None. zThe PyTorch API of SparseSemiStructuredTensor is in prototype stage and will change in the near future. Please open a Github issue for features requests and see our documentation on the torch.sparse module for further information about the project. stacklevelTz3At least one of packed or packed_t must be provided)devicedtypelayoutr*)rwarningswarn UserWarning_load_dispatch_tabletorch_dynamoallow_in_graph ValueErrorTensor_make_wrapper_subclassr/r0r1r"r#r$r%r&r'r() clsr)r"r#r$r%r&r'r(r*previous_tensortensors [/var/www/html/ia-enhanced/venv/lib/python3.12/site-packages/torch/sparse/semi_structured.py__new__z"SparseSemiStructuredTensor.__new__OsN++ MMH  ,0C (  $ $ & MM ( ( -  $O  !&ORS S44  "))!''"))' 5   " -H*+D(#4  returnc|t|ds td|jjd|jdS)Nr)ztensor has no shape attributez(shape=))hasattrAssertionError __class____name__r))selfs r?__repr__z#SparseSemiStructuredTensor.__repr__s:tW% !@A A..))*'$**Q??rActtfdj}jjj j f}||fS)Nc t|duSN)getattr)xrIs r?z?SparseSemiStructuredTensor.__tensor_flatten__..sWT1-T9rA)listfilter __slots__r)r'r(r*)rI inner_tensors tensor_metas` r?__tensor_flatten__z-SparseSemiStructuredTensor.__tensor_flatten__sV 94>> J  JJ  * *  " "     k))rArUc |\}}}}|||jdd|jdd|jdd|jdd|jdd||| S)Nr"r#r$r%r& r)r"r#r$r%r&r'r(r*)get) r<rTrU outer_size outer_strider)r'r(r*s r?__tensor_unflatten__z/SparseSemiStructuredTensor.__tensor_unflatten__sNYJ(*;] $$Xt4""640"&&z48 $$Xt4(5(9(9-t)'@/'  rAc|j|jvr%t|jd|jd|j|j||||S)NzI only supports a specific set of operations, can't perform requested op (rD)_overloadpacketr!NotImplementedErrorrH)r<functypesargskwargss r?__torch_dispatch__z-SparseSemiStructuredTensor.__torch_dispatch__sh   s':': :%<<.!//3}}oQ@ 9s""4#7#78udFSSrAct|dd~tjjjt tjjj ttjjjttjjjttjjjttjjjttjjjt tjjj"t$tjjj&t$tjjj(t*tjjj,t.tjjj0t2tjjj4t6tjjj8t:tjjj<t>i|_ ||j@jC|yyy)zT Loads the op overload sparse dispatch table for the current class. r!N)"rNr6opsatenvaluesrindicesr is_same_sizerdetach_detachr trviewrmmrmatmuladdmmr linearr _to_copyr _scaled_mmrcloner torr!update)r<custom_dispatch_tables r?r5z/SparseSemiStructuredTensor._load_dispatch_tablesm 3)4 0 8 %%'9 &&(; ++-@ &&(; %%'9   - ##%5 !!> %%~ $$&7 %%'9 '')< ))+@ $$&7 !!>#C "%0##**+@A1% 9rAoriginal_tensorc \|jstd|jd|jdk7rtd|jd|j s td|j |j vrtd|j d|d |j\}}|j |j j}|j |j j}||ks||zs ||ks||zrtd |jd |d |d y)z_ Assert that the given tensor is valid for semi-structured sparse compression. zError original_tensor.device= z= is not supported! Only CUDA tensors are currently supported.r,zError original_tensor.dim = z; is not supported! Only 2d tensors are currently supported.zXError original_tensor is not contiguous!Only contiguous tensors are currently supported.zError original_tensor.dtype z is not a supported dtype for !zError original_tensor.shape zS is not supported! Both dimensions must be larger or equal than and a multiple of (z, rDN) is_cuda RuntimeErrorr/dim is_contiguousr0rr)sparse_min_rowssparse_min_cols)r<rymnmin_rowsmin_colss r? _validate_device_dim_dtype_shapez;SparseSemiStructuredTensor._validate_device_dim_dtype_shapesr &&01G1G0HI==     A %./B/B/D.EF;;  ,,.C   (D(D D./D/D.EEcdgchhij  $$1//0E0EFVV//0E0EFVV x<1x<1x<1x<./D/D.EFSS[R\\^_g^hhik >> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_CUDA) >>> A = torch.Tensor([0, 0, 1, 1]).tile((128, 32)).half().cuda() tensor([[0., 0., 1., ..., 0., 1., 1.], [0., 0., 1., ..., 0., 1., 1.], [0., 0., 1., ..., 0., 1., 1.], ..., [0., 0., 1., ..., 0., 1., 1.], [0., 0., 1., ..., 0., 1., 1.], [0., 0., 1., ..., 0., 1., 1.]], device='cuda:0', dtype=torch.float16) >>> A_sparse = to_sparse_semi_structured(A) SparseSemiStructuredTensor(shape=torch.Size([128, 128])) >>> A_sparse.values() tensor([[1., 1., 1., ..., 1., 1., 1.], [1., 1., 1., ..., 1., 1., 1.], [1., 1., 1., ..., 1., 1., 1.], ..., [1., 1., 1., ..., 1., 1., 1.], [1., 1., 1., ..., 1., 1., 1.], [1., 1., 1., ..., 1., 1., 1.]], device='cuda:0', dtype=torch.float16), >>> A_sparse.indices() tensor([[-4370, -4370, -4370, ..., -4370, -4370, -4370], [-4370, -4370, -4370, ..., -4370, -4370, -4370], [-4370, -4370, -4370, ..., -4370, -4370, -4370], ..., [-4370, -4370, -4370, ..., -4370, -4370, -4370], [-4370, -4370, -4370, ..., -4370, -4370, -4370], [-4370, -4370, -4370, ..., -4370, -4370, -4370]], device='cuda:0', dtype=torch.int16)) zSetting transpose from `to_sparse_semi_structured` is deprecated and will be removed in a future release. `SparseSemiStructuredTensor` only support contiguous input tensors.r,r-)r) r2r3 FutureWarningrrr6sparserrr)ryrrSPARSE_SUBCLASSs r?rr5sfh  R    & 4 4  66 \\ > >  % %of % EErAc eZdZdZdZej eddddejeddddejeddddejeddddiZ e e jfd ejd ed dfd Zfd Ze dd ejd dfdZddddejdejdzded ejfdZxZS)ra This class implements semi-structured sparsity for the CUTLASS backend. In this implementation, the specified elements and metadata are stored separately, in packed and meta respectively. When _FORCE_CUTLASS is set, or when cuSPARSELt is not available, this subclass calls into _sparse_semi_structured_(mm|addmm) and sparse_semi_structured_from_dense for conversion to the compressed format. cutlass @ryrrBc |j|t|\}}||j||ddd|jS)Nr"r#r$r%r&r*)rrr)r*)r<ryrsparse_tensor_cutlassmeta_tensor_cutlasss r?rz,SparseSemiStructuredTensorCUTLASS.from_densesW ,,_= 6o F !   ! !($(,)77  rAc|j |j td|jjdk(r t |j|jSt |S)Nz meta and packed must not be Noner,)r#r"rFndimrsuperr)rIrGs r?rz*SparseSemiStructuredTensorCUTLASS.to_densesc 99  3 !CD D yy~~" 4    !#  rArc ptj||d\}}}}}||j|||||dS)a~ This function takes in a unpruned dense tensor and runs a (branchless) static sort across a 4x4 tile. It greedily picks the largest values in the tile, upholding the 2:4 sparsity constraint across both rows and columns. The algorithm used to prune the matrix is implemented in `_sparse_semi_structured_tile`. Then it creates the packed and meta tensors for the compressed sparse representation of the pruned dense tensor. It also calculates the packed_t and meta_t tensors for the compressed sparse representation of the transposed pruned dense tensor. Since we cannot transpose the compressed representations, we store both for the fw/bw pass respectively. Finally, this function also computes a compressed swizzled bitmask that encodes the sparsity pattern This can be used in the backward pass to mask the gradients. [9 1 7 4] [9 0 7 0] [1 2 3 0] [0 2 0 0] [8 3 5 4] -> prune 4x4 tile -> [8 0 0 4] -> pack to CUTLASS semi-structured -> packed [1 2 6 2] [0 0 6 2] -> metadata -> pack to transposed CUTLASS -> packed_t semi-structured representation -> metadata_t -> compute swizzled bitmask -> compressed_swizzled_bitmask The equivalent PyTorch code to create the same five outputs from the dense tensor can be found below: ``` from torch.sparse import SparseSemiStructuredTensorCUTLASS from torch.sparse._semi_structured_conversions import ( _sparse_semi_structured_tile, _compute_compressed_swizzled_bitmask, ) pruned = _sparse_semi_structured_tile(dense) packed_cutlass, meta_cutlass = sparse_semi_structured_from_dense_cutlass(pruned) packed_t_cutlass, meta_t_cutlass = sparse_semi_structured_from_dense_cutlass( pruned.t().contiguous() ) bitmask = _compute_compressed_swizzled_bitmask(pruned) SparseSemiStructuredTensorCUTLASS( dense.shape, packed_cutlass, meta_cutlass, packed_t_cutlass, meta_t_cutlass, bitmask, ) ``` T algorithm use_cutlassFr)r6_sparse_semi_structured_tiler)r<ryrr"r#r$r%r&s r?prune_dense_static_sortz9SparseSemiStructuredTensorCUTLASS.prune_dense_static_sortsXz  . . yd      '   ! !(C  rANFrshould_transpose_denserrrc t|tr td|jj}|j dk7s|j dk7rt d|d|j |jt d|dt|j|j}tjjj||j|j||j d|j"|j$|S)NZ`SparseSemiStructuredTensor @ SparseSemiStructuredTensor` is not supported by the hardwarer,`)` matmul: Broadcasting is not implemented$` matmul: operation is not supportedr) isinstancerr9rGrHrr_r"r#_ensure_cutlass_mm_registeredrr0r6rfsemi_structured cutlass_mmr)dense_min_rowsdense_min_cols)rIrrrrccls_name constraintss r?rz%SparseSemiStructuredTensorCUTLASS._mms a3 4l >>** 99>QVVq[%H:FG  ;; $))"3%H:AB  * +77@K99,,77   1 ****&  rA)rHrrrr r6int8rfloat16bfloat16float32rrrrr:rrrrrr __classcell__)rGs@r?rr|s# G 22sBC 5b"aC 6r2q!D 5b"aC  1@@    -   *  68I #llI %I I ^%)', ! <<!llT! ! !% ! !rArc eZdZdZdZej eddddejeddddejeddddejeddddiZ e e jfdejdeddfd Ze ddejdd fd Zd d ddejdejd zdedejfdZy )ra The cuSPARSELt backend expects the specified elements and the metadata to be stored in a single tensor: packed = [ specified elements of original tensor | metadata ] For an original tensor of size (m, k) we expect the first m * k // 2 elements to be the kept elements The rest of the tensor is metadata. Since there is only one tensor, we only use the packed and packed_t attributes respectively. cuSPARSELt also supports transposition fusion, which is necessary for performant 2:4 sparse training, as well as specifying alg_id, a config that affects the performance of the matmul depending on matmul sizes. cusparseltrrrryrrBc |j|||jtj|ddddtj ||j  S)NrX)rr)r6_cslt_compressrrr*rs r?rz/SparseSemiStructuredTensorCUSPARSELT.from_dense6sW ,,_=!''''8(,&@&P&P$)77  rArc tj||d\}}}}}|j|jdd}|j|jdd}||j|||||dS)a= This function does the same thing as described in SparseSemiStructuredCUTLASS, but uses the cuSPARSELt metadata layout and sparse matmul. The only functional difference is that cuSPARSELt stores `metadata` and `packed` together into a single tensor. [9 1 7 4] [9 0 7 0] [1 2 3 0] [0 2 0 0] [8 3 5 4] -> prune 4x4 tile -> [8 0 0 4] -> pack to cuSPARSELT semi-structured -> packed [1 2 6 2] [0 0 6 2] -> pack to transposed cuSPARSELt -> packed_t semi-structured representation -> compute swizzled bitmask -> compressed_swizzled_bitmask The equivalent PyTorch code to create the same three outputs from the dense tensor can be found below: ``` from torch.sparse import SparseSemiStructuredTensorCUSPARSELT from torch.sparse._semi_structured_conversions import ( _sparse_semi_structured_tile, _compute_compressed_swizzled_bitmask, ) pruned = _sparse_semi_structured_tile(dense) packed_cusparselt = torch._cslt_compress(pruned) packed_t_cusparselt = torch._cslt_compress(pruned.t().contiguous()) bitmask = _compute_compressed_swizzled_bitmask(pruned) SparseSemiStructuredTensorCUSPARSELT( dense.shape, packed_cutlass, None, packed_t_cutlass, None, bitmask ) ``` Frrrr)r6rrnr)rs r?rz  ! !(C  rANFrrrrc <t|tr td|jdk7s|jdk7r#t d|j j d|j|jk7rit d|j j dt|jdt|jd|jd|jd ||j|jk7rit d|j j dt|jdt|jd |jd |jd |jtjk(r\t d|j j dt|jdt|jd |jd |j#t d|j j dt|j|j}tjj j#||j||jd|j$|j&|j(|j*| S)Nrr,rrz` matmul: trying to do `A=z @ B=z`, with A.dtype=z and B.dtype=zH. This operation is only supported when A and B have the same data type.z + C`, with A.dtype=B.dtype=z and C.dtype=zK. This operation is only supported when A, B and C have the same data type.z`, with A.dtype=B.dtype=zO. mm is not supported for float8_e4m3fn, please use `torch._scaled_mm` instead.rr)rrr9rr_rGrHr0rr)r6 float8_e4m3fnr" _ensure_cusparselt_mm_registeredrrfr cusparselt_mmrrr'r()rIrrrrcrs r?rz(SparseSemiStructuredTensorCUSPARSELT._mmsw a3 4l  99>QVVq[%DNN++,,UV  77djj %DNN++,,FuTZZGXFYY^_defelel_m^no $ |= BYY    djj 8%DNN++,,FuTZZGXFYY^_defelel_m^no((, |= J\\  ::,, ,%DNN++,,FuTZZGXFYY^_defelel_m^no((, |4``  ;; %DNN++,,PQ  - .77@K99,,::  1 ****..&&&  rAr)rHrrrr r6rrrrrrrrrr:rrrrrrrAr?rr"s G ;BBK 22r2rB 5b"aC 6r2q!D  1@@    0   &68> #ll> %> > H%)', 4 <<4llT! 4 !% 4 4rArctrydaddlm}|dddtjd tjd tjd tjdzd t d t dt dt dtjfd}|jdtjd tjd tjd tjdzd t d t dt dt dtjfd}y)zLazily register the cutlass_mm custom op. Registration is deferred to avoid importing torch.library at module load time, since torch.sparse is imported early during ``import torch``. NTr custom_opzsemi_structured::cutlass_mmr mutates_argsdenser"r#r out_featuresrrrrBc|j\}} | |z} | |z} | dk7xs| dk7} |} | r.tjjj |d| d| f} |r| j n| }|tj |||}ntj||||}| r:|r|n| }|d|jdd|jtjS|jS)Nrr memory_format) r)r6nn functionalpadrm_sparse_semi_structured_mm_sparse_semi_structured_addmmnarrowrucontiguous_format contiguous)rr"r#rrrrrrrto_pad_mto_pad_nneed_pad dense_paddedmm_inputresout_colss r?rz1_ensure_cutlass_mm_registered..cutlass_mms{{1B(?B(?q=1HM  88..2251h8:TUL'=<>>#< <2264JC55dFD(SC 2qHM\"1h'U%<%<=  ~~rAtranspose_densec|r|jdn|jd}tj|||j|jSNrrrr)r6emptyr0r/) rr"r#rrrrrrs r?_cutlass_mm_fakez7_ensure_cutlass_mm_registered.._cutlass_mm_fakesB&55;;q>%++a.{{  ++<<   rA)_cutlass_mm_registered torch.libraryrr6r:rr register_fake)rrrs r?rrs%!',2> ||   ll llT!       !%   ? > ||   ll llT!           rAc trydaddlm}|dd ddtjd tjd tjdzd t d t d t dt dt dt dtjfd}|jdtjd tjd tjdzd t d t d t dt dt dt dtjfd}y)z,Lazily register the cusparselt_mm custom op.NTrrzsemi_structured::cusparselt_mmrrrr"rrrrfuse_transposerrrBc |j\} } | |z} | |z} | dk7xs| dk7} |}| r.tjjj |d| d| f}|r|j n|}tj |||||}|r|j }| r7|r| n| }|jdd|jtjS|jS)Nr)rtranspose_resultrrr) r)r6rrrrm_cslt_sparse_mmrrurr)rr"rrrrrrrrrrrrrrrrs r?rz7_ensure_cusparselt_mm_registered..cusparselt_mms{{1B(?B(?q=1HM  88..2251h8:TUL'=<>>#<##  +   %%'C 2qH::aH-33#554 ~~rAc |r|jdn|jd} tj|| |j|jSrr) rr"rrrrrrrrs r?_cusparselt_mm_fakez=_ensure_cusparselt_mm_registered.._cusparselt_mm_fake6sB&<5;;q>Q{{  ++<<   rA)F)_cusparselt_mm_registeredrrr6r:rrr)rrr s r?rr s8! $'/bA(-! ||!  ! llT!!  !  !  ! ! ! !%!  ! B! F   ||   llT!         !%   ! rA)'r2 collectionsrcollections.abcrtypingrr6)torch.sparse._semi_structured_conversionsrr!torch.sparse._semi_structured_opsrr r r r r rrrrrrr__all__rr:rrrrrrrrrr rrrAr?rs"$ " ",$C" H"H"Z,<<DF\\DFDF DF DFNc(BcL]+E]@> B"? rA