a df;@sddlZddlmZddlmZmZmZmZddlm Z ddl Z ddl Z ddl Z ddl Z ddlZ ddlZ ddlZ ddlZ ddlmZddlmZddlmZeddd Zedd d ZGd d d e jjZGddde jjZddZddddZ GdddeZ!dS)N) nullcontext)AnyCallableDictSequence)warn)NvfuserPrimOperatorSupport)torch_function_passthrough)get_isolated_graphmodulec*Csttjftjtjjftjjtjjjftjtjjftjtjjftjtjjfg}tjjtjj tjj tjj tjj tjj tjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjjtjj tjj tjj!tjj!tjj"tjj"tjj#tjj#tjj$tjj$tjj%tjj%tjj&tj'j(tjj)tj'j)i}|D]0\}}|j*D]}|j+,|||j+,|<q^qPt-tjD].}|tjj*vrtjj+,||t.tj|<qtjj/j*D]4}t.tj|dpt.t|}tjj/j+,|||<q|S)z} Mapping of torch API functions to torch._refs functions. E.g. torch_to_refs_map()[torch.add] == torch._refs.add N)0torchZ_refsnnZ functionalZspecialZfftZlinalgZTensor __invert__Z bitwise_not__xor__Z bitwise_xor__and__Z bitwise_and__or__Z bitwise_or__eq__eq__rsub__Zrsub __rtruediv__Zrtruediv __floordiv__Z floor_divide __rfloordiv__Z rfloordiv__pow__pow__rpow__ZrpowZ new_emptyZnew_full new_zerosZnew_onesZfill_Zzero_toZ sum_to_sizeZcopy__primsZcopy_toresize__all____dict__getdirgetattrZ _conversions)modulesrZ mod_torchZmod_refssZ tensor_attrr&]/var/www/html/stable-diffusion-webui/venv/lib/python3.9/site-packages/torch/_prims/context.pytorch_to_refs_mapsN                            r(cCsddtjjDS)zJ Set of all prim functions, e.g., torch._prims.add in all_prims() cSsh|]}tjj|qSr&)r rrr ).0r%r&r&r' Rzall_prims..)r rrr&r&r&r' all_primsMsr,c@s8eZdZdZddddZd eeeeeddd Z dS) NvfuserPrimsModea` Switches the interpretation of torch.ops.prims.* functions to use nvFuser's prims in torch.ops.nvprims.* >>> # xdoctest: +SKIP("undefined vars") >>> with NvfuserPrimsMode(): ... torch.ops.prims.add(x, y) # calls torch.ops.nvprims.add(x, y) By default, this context manager will fall back on the torch.ops.prims* if the nvprim does not exist. It's possible to skip certain prims by passing their names to the skip_ops argument. skip_ops is expected to be a sequence of strings, e.g., ["prims.add.default"] In order to check the expected name of a prim, one can use the `torch.overrides.resolve_name`. >>> # xdoctest: +SKIP("undefined vars") >>> with NvfuserPrimsMode(skips_ops=("prims.add.default")): ... torch.ops.prims.add.default(x, y) # does not call torch.ops.nvprims.add.default(x, y) r&skip_opscCs ||_dSNr.)selfr/r&r&r'__init__jszNvfuserPrimsMode.__init__N orig_functypesargskwargscCs|dur i}tj||jvr,||i|St|tjjtjjfrt| dd}t| dd}|dkrt tj j |d}|dur||i|S||i|S)N.rZprims) r overrides resolve_namer/ isinstance_ops OpOverloadOpOverloadPacketstrsplitr"opsnvprims)r1r4r5r6r7 namespacenameZnvfuncr&r&r'__torch_function__msz#NvfuserPrimsMode.__torch_function__)r&N) __name__ __module__ __qualname____doc__r2rrrrrFr&r&r&r'r-Usr-c@s>eZdZdZdddefddZd eeeee d d d Z dS) TorchRefsModeaB Switches the interpretation of torch.* functions and Tensor methods to use PrimTorch refs in torch._refs. (Direct calls to _refs are unaffected.) >>> # xdoctest: +SKIP >>> with TorchRefsMode(): ... torch.add(x, y) # calls torch._refs.add(x, y) By default, this context manager will fall back on the torch.* if the ref does not exist; set strict=True to error if this occurs. If the ref exists we still would like to fall back on the torch.* sometimes, this behavior can be customized by passing a function to should_fallback_fn. FcGsdS)NFr&)_r&r&r'r+zTorchRefsMode.cCs||_||_||_dSr0strictshould_fallback_fnprims_mode_cls)r1rOrPrQr&r&r'r2szTorchRefsMode.__init__r&Nr3cCs|dur i}|tvs|tvrT|||i|WdS1sJ0Yt}||d}|durt|tjjrtj j |d}|dur| |||||r||i|S|||i|WdS1s0Y|j rt dtj|||i|S)Nzno _refs support for )r r,rQr(r r<r r=r>_decompdecomposition_tablerPrO RuntimeErrorr:r;)r1r4r5r6r7mappingfuncr&r&r'rFs& ,  ,z TorchRefsMode.__torch_function__)r&N) rGrHrIrJrr2rrrrrFr&r&r&r'rKs rKcCs|jdkot|jddduS)N call_functionZ impl_nvfuser)opr"target)noder&r&r'_is_node_supported_nvfusers r[r&r.c stj||vrdS|nzt|||}WnLtyv}z4td|jdt|WYd}~WddSd}~00Wdn1s0Ytt dd|j j }t fdd|D} | S) a This function traces the `func` under `torch_function_mode` and checks if any of the traced nodes are not supported by nvFuser. If so, we should fallback to the original function. `skip_ops` argument is expected to be a list of strings of function names that would match with `torch.overrides.resolve_name`. Args: torch_function_mode: The torch_function_mode context manager. orig_func: The original function, its name will be used to check if it should be skipped. func: The function to be traced. args: The args to be passed to the function. kwargs: The kwargs to be passed to the function. Keyword args: skip_ops: A list of ops to skip when checking if the function is supported. Tz2get_isolated_graphmodule failed on decomposition: z with error message: NcSs |jdkS)NrW)rX)nr&r&r'rMr+z._is_func_unsupported_nvfuser..c3s|]}d| VqdSr0)Zis_node_supported)r)rZZ supported_opsr&r' sz/_is_func_unsupported_nvfuser..) r r:r;r ExceptionrrGr@rfiltergraphnodesany) Ztorch_function_moder4rVr6r7r/ZgmeZcall_function_nodesZany_unsupportedr&r]r'_is_func_unsupported_nvfusers,B recsxeZdZddfdd ZddZddZd d Zd d Zd dZddZ ddZ de e e e edfdd ZZS)TorchRefsNvfuserCapabilityModer&r.csDd}t|||_tjdtjtt||dtjt|dddS)N)zaten._log_softmax.defaultz'aten._log_softmax_backward_data.defaultzaten.expand.defaultFr.rN)tupler/superr2 functoolspartialrer-)r1r/Zaten_ops_to_skip __class__r&r'r2s  z'TorchRefsNvfuserCapabilityMode.__init__c Csbtjj||||||||\} } } |r>| | | |jdtjdfS| |d|d|jdtjdfS)N)r)Zdtype)r rBrCnative_batch_normrZuint8) r1inputweightZbias running_mean running_varZtrainingZexponential_average_factorepsilonabcr&r&r'_cudnn_batch_norms"  z0TorchRefsNvfuserCapabilityMode._cudnn_batch_normc Cs2tjjtjjjj} | |||||||d|gd S)NT)TTT)r rRrSrBatenZnative_batch_norm_backwarddefault) r1rnZ grad_outputrorprqZ save_meanZsave_varrrZ reserveSpacerVr&r&r'_cudnn_batch_norm_backward0s  z9TorchRefsNvfuserCapabilityMode._cudnn_batch_norm_backwardcCs2dtj|kp0t|tjjtjjfo0dt|vS)Nztorch.var_meanz aten.var_mean)r r:r;r<r=r>r?r@r1rVr&r&r' _is_var_meanLs z+TorchRefsNvfuserCapabilityMode._is_var_meancCs"hdt|j}tj||vS)N>ztorch.view_copyztorch.Tensor.reshapeztorch.Tensor.viewzaten.view_copy.defaultz torch.reshapezaten._unsafe_view.defaultzaten.view.default)setr/r r:r;)r1rVZ allowed_opsr&r&r'_is_view_or_reshapeRs z2TorchRefsNvfuserCapabilityMode._is_view_or_reshapecCs.dtj|kp,|tjjjjkp,|tjjjkS)Nztorch.native_batch_norm)r r:r;rBrwrmrxrzr&r&r'_is_native_batch_norm^s z4TorchRefsNvfuserCapabilityMode._is_native_batch_normcCs2dtj|kp,|tjjjkp,|tjjjjk}|S)Nztorch.rand_like)r r:r;rBrw rand_likerxr1rVresultr&r&r' _is_rand_likedsz,TorchRefsNvfuserCapabilityMode._is_rand_likecCs.dtj|kp(|tjjjtjjjjfv}|S)Nz torch.full)r r:r;rBrwfullnamesrr&r&r'_is_fulljs z'TorchRefsNvfuserCapabilityMode._is_fullNr3cs|dur i}||r*tjjj|i|S|tjjjjksH|tjjjkr|| |j|i|WdS1sr0Y|tjjj jks|tjjj kr| |j |i|WdS1s0Y| |r|^}}tj j |dd}t|dkrtdtjj||S|tjjjjkrZ|\}}}t|dkrJtdtjj||S||rztjjj|i|S||rt|dkrtdtjjj|S||rtjjj|i|St||||S)NF)validaterzview has ignored kwargs!zrand_like has ignored kwargs!)r{r rBrCZvar_meanrwZcudnn_batch_normrxrvZcudnn_batch_norm_backwardryr}Z _prims_commonZextract_shape_from_varargslenrviewZ_reshape_aliasr~rmrrrrrhrF)r1r4r5r6r7rsshapeZstriderkr&r'rFtsL  . .     z1TorchRefsNvfuserCapabilityMode.__torch_function__)r&N)rGrHrIr2rvryr{r}r~rrrrrrrF __classcell__r&r&rkr'rfs  rf)"ri contextlibrtypingrrrrwarningsrr Z torch._decompZ torch._primsZ torch._refsZtorch._refs.nnZtorch._refs.nn.functionalZtorch._refs.specialZtorch.overridesZtorch._prims.nvfuser_executorrZtorch._prims_commonr Z"torch.fx.experimental.proxy_tensorr lru_cacher(r,r:ZTorchFunctionModer-rKr[rerfr&r&r&r's0      6 1A /