MCPcopy Create free account

hub / github.com/alibaba/BladeDISC / functions

Functions6,088 in github.com/alibaba/BladeDISC

↓ 8 callersMethodget_node
(self, graph_or_func, op)
tao/tao_bridge/test/gpu/mlir/base.py:41
↓ 8 callersMethodisShapeEqual
tao_compiler/mlir/disc/transforms/shape_utils.cc:838
↓ 8 callersMethodis_empty
Return whether the tensor is empty
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:614
↓ 8 callersMethodmake_grouped_weights
no data copy
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:672
↓ 8 callersFunctionmayConvertToIndexType
tao_compiler/mlir/disc/transforms/codegen_utils.cc:143
↓ 8 callersMethodmodify_graph
(self, passes: Iterable[GraphModPass])
tools/torch_quant/torch_quant/graph.py:137
↓ 8 callersMethodnewConstantSymbolicDim
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:109
↓ 8 callersMethodnodes_by_module_type
(self, module_types: Iterable[Type[nn.Module]])
tools/torch_quant/torch_quant/graph.py:141
↓ 8 callersMethodnum_nodes
Get total number of nodes in graph.
tensorflow_blade/tf_blade/util/simple_graph.py:109
↓ 8 callersMethodprepare
tao_compiler/mlir/xla/ral/context/common_context_impl_acl.cc:290
↓ 8 callersMethodreuse_packed_weight
tao_compiler/mlir/xla/ral/context/common_context_impl_acl.cc:361
↓ 8 callersFunctionrow_to_col
(input_: Tensor, dtype=torch.float)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/distributed/comm.py:103
↓ 8 callersFunctionstrideAllOnes
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:240
↓ 8 callersFunctiontensor_scale_mask
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/utils.hpp:264
↓ 8 callersFunctionwhich
Same as `which` command of bash.
scripts/python/common_setup.py:174
↓ 8 callersFunctionwriteToBinaryStream
tao_compiler/mlir/xla/ral/ral_metadata.cc:47
↓ 7 callersFunctionBackwardsConstAnalysis
Backwards dataflow analysis that finds arguments to a graph that must be compile-time constants.
tao/tao_bridge/tf/const_analysis.cc:222
↓ 7 callersFunctionBranch_Name
tao/tao_bridge/passes/functionalize_cond.cc:53
↓ 7 callersFunctionGetDeviceAddress
tao_compiler/mlir/xla/ral/context/stream_executor_based_impl.cc:629
↓ 7 callersFunctionGetMarkForCompilationPassFlags
tao/tao_bridge/tf/flags.cc:234
↓ 7 callersFunctionMLIRType2CUDATypeStr
tao_compiler/mlir/disc/utils/source_emitter.cc:192
↓ 7 callersMethodNewNode
tao/tao_bridge/tf/graphcycles.cc:110
↓ 7 callersFunctionNodesToString
tao/tao_bridge/passes/functionalize_control_flow_util.h:50
↓ 7 callersMethod__init__
(self, chunk_traced_module)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/utils.py:139
↓ 7 callersFunction_create_graph_def
px -> Add --> Mul -> Relu -> ABS -> Add -> Sqrt / \_/ / \ # noqa: W605 py -
tensorflow_blade/tests/util/simple_graph_test.py:54
↓ 7 callersMethod_test_batchnorm
(self, batchnorm_func, inp_test_data=None)
pytorch_blade/tests/disc/ops/test_batchnorm.py:19
↓ 7 callersMethod_test_conv
(self, conv_func, inp_test_data=None)
pytorch_blade/tests/disc/ops/test_conv_ops.py:20
↓ 7 callersFunctionaddFusionTag
Attaches a new tag to the fusion op. Here different tags is mapping to different variants of the fusion op.
tao_compiler/mlir/disc/transforms/fusion_utils.cc:260
↓ 7 callersMethodallocate
tao_compiler/mlir/xla/ral/context/common_context_impl_acl.cc:110
↓ 7 callersMethodasCUStream
tao_compiler/mlir/xla/ral/device/gpu/gpu_driver.cc:238
↓ 7 callersFunctioncol_to_row
(input_: Tensor, dtype=torch.float)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/distributed/comm.py:94
↓ 7 callersMethoddominates
tao_compiler/mlir/disc/transforms/disc_transpose_simplifier.cc:419
↓ 7 callersFunctionemitNumElementsComputation
tao_compiler/mlir/disc/transforms/codegen_utils.cc:92
↓ 7 callersFunctionemitWidthAdaptShuffle
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:907
↓ 7 callersFunctionfmap
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/utils.hpp:65
↓ 7 callersMethodgetCanonicalizationPatterns
tao_compiler/mlir/disc/tools/disc-transform/LinalgExt/LinalgExtOps.cc:651
↓ 7 callersMethodgetOrCreateResource
tao_compiler/mlir/xla/ral/ral_context.cc:209
↓ 7 callersFunctiongetOrParsePDLAttr
tao_compiler/mlir/xla/ral/context/pdll_util.h:213
↓ 7 callersFunctiongetSymDimSize
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:71
↓ 7 callersFunctiongetThreadPerBlock
tao_compiler/mlir/disc/transforms/codegen_utils.cc:65
↓ 7 callersMethodget_current_context
Return the deepest context on the stack.
pytorch_blade/torch_blade/config.py:118
↓ 7 callersMethodhas_zero_point
Return whether the param has a zero_point
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:901
↓ 7 callersMethodisFusible
tao_compiler/mlir/disc/transforms/fusion_utils.cc:1123
↓ 7 callersFunctionisOnGpu
tao_compiler/mlir/disc/transforms/fusion_utils.cc:153
↓ 7 callersFunctionisSameUnderlineBuffer
tao_compiler/mlir/disc/disc_util.cc:81
↓ 7 callersFunctionisWeightPrePackingForMatMulEnabled
tao_compiler/mlir/xla/ral/context/common_context_impl_mkldnn.cc:140
↓ 7 callersMethodis_reachable
Check if any nodes with index in `target_idx` can be reached from `src_idx` node. Parameters ---------- src_idx : int
tensorflow_blade/tf_blade/util/simple_graph.py:214
↓ 7 callersFunctionloadOrCreateGlobalString
Creates a global const string op named `name` using the value if not exists and returns the Loaded value of this global op.
tao_compiler/mlir/disc/transforms/disc_to_llvm.cc:183
↓ 7 callersMethodmatch
tao_compiler/mlir/disc/transforms/disc_lower_gpu_ops_common.h:41
↓ 7 callersFunctionop_scale_mask
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/utils.hpp:262
↓ 7 callersFunctionoptimize_module
( model, dummy_inputs, optimize_config=None, enable_disc=True, enable_amp=True, en
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/optimize.py:101
↓ 7 callersFunctionregisterBufferizableOpInterfaceExternalModels
tao_compiler/mlir/disc/tools/disc-transform/LinalgExt/LinalgExtOps.cc:1013
↓ 7 callersFunctiontoIValue
pytorch_blade/pytorch_blade/compiler/jit/torch/pybind_utils.cpp:35
↓ 7 callersFunctiontoggle_observer
(root: nn.Module, *, observe: bool, fake_quant: bool)
tools/torch_quant/torch_quant/observer.py:173
↓ 6 callersMethodAllNodesInPostOrder
tao_compiler/mlir/disc/utils/cycle_detector.cc:326
↓ 6 callersMethodCondStateToString
tao/tao_bridge/passes/functionalize_cond.cc:227
↓ 6 callersMethodContractEdge
tao_compiler/mlir/disc/utils/cycle_detector.cc:273
↓ 6 callersMethodErase
Removes `value` from the set. Assumes `value` is already present in the set.
tao_compiler/mlir/disc/utils/cycle_detector.h:59
↓ 6 callersMethodGet
tao_compiler/mlir/xla/ral/ral_context.cc:44
↓ 6 callersFunctionGetAllPossibleUsedValues
tao_compiler/mlir/disc/disc_util.cc:189
↓ 6 callersFunctionGetResourceOpInfoForOp
tao/tao_bridge/tf/resource_operation_table.cc:120
↓ 6 callersFunctionHasOpType
tao/tao_bridge/tao_util.cc:25
↓ 6 callersFunctionIsReachable
Return whether "to" is reachable from "from".
tao/tao_bridge/tf/graphcycles_test.cc:55
↓ 6 callersMethodIsReachable
tao_compiler/mlir/disc/utils/cycle_detector.cc:254
↓ 6 callersMethodMerge
tao/tao_bridge/tf/union_find.h:53
↓ 6 callersFunctionPartiallyDecluster
tao/tao_bridge/passes/tao_partially_decluster_pass_test.cc:102
↓ 6 callersFunctionRandomNode
Returns the index of a randomly chosen node in *nodes. Requires *nodes be non-empty.
tao/tao_bridge/tf/graphcycles_test.cc:188
↓ 6 callersFunctionRatioToString
tao/tao_bridge/passes/tao_mark_for_compilation_pass.cc:1818
↓ 6 callersMethodSetProgram
tao/tao_bridge/tf/subprocess.cc:84
↓ 6 callersFunctionVectorString
tao/tao_bridge/tf/util.h:356
↓ 6 callersFunctionVerify
tao_compiler/mlir/disc/IR/disc_shape_ops.cc:28
↓ 6 callersMethod__init__
(self)
pytorch_blade/tests/disc/pdl/test_torchscipt_to_mhlo/test_quantization.py:33
↓ 6 callersFunction_create_graph_def_front_circle
-> Relu - / \ # noqa: W605 px -> Add -> Mul -> ABS Mul -> Relu
tensorflow_blade/tests/util/simple_graph_test.py:99
↓ 6 callersFunction_locate_parent
(root: nn.Module, full_path: str)
tools/torch_quant/torch_quant/graph.py:59
↓ 6 callersMethod_test_cmp_func
(self, torch_func)
pytorch_blade/tests/disc/ops/test_binary_ops.py:235
↓ 6 callersMethod_test_factory_like
(self, func, d0: int, d1: int, d_or_val: int, numeric_checkable=True, expect_str=None)
pytorch_blade/tests/disc/ops/test_factory_like.py:21
↓ 6 callersMethod_test_model
(self, func_name, config_path, inference_func, batch=1)
pytorch_blade/benchmark/detectron2/run_blade.py:221
↓ 6 callersMethodappend
(self, name, secs)
scripts/python/common_setup.py:47
↓ 6 callersFunctionbuildDISCBufferize
tao_compiler/mlir/disc/transforms/disc_transform_schedule.cc:213
↓ 6 callersFunctionbuildDISCFuseIntoContainingOp
tao_compiler/mlir/disc/transforms/disc_transform_schedule.cc:294
↓ 6 callersFunctionbuildGetParentForOp
tao_compiler/mlir/disc/transforms/disc_transform_schedule.cc:186
↓ 6 callersFunctionbuildPadOp
tao_compiler/mlir/disc/transforms/disc_transform_schedule.cc:169
↓ 6 callersFunctioncopy_and_replace
(root: nn.Module, trace_mapping: Dict[str, TracePair])
tools/torch_quant/torch_quant/module.py:155
↓ 6 callersFunctioncwd
Change the current working directory to `path` to do somthing and then recover the current cwd when it's done.
pytorch_blade/common_setup.py:185
↓ 6 callersFunctiondetermineListSize
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:850
↓ 6 callersMethodencode
encodes input (uint4) into output (unsigned char). Assumes len is a multiple of 4.
tao_compiler/mlir/xla/ral/ral_md5.cc:157
↓ 6 callersFunctiongetDeviceFromValue
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:83
↓ 6 callersFunctiongetDimSize
Returns size of dimension at the specified index, if ranked tensor. Otherwise, returns -1. Aborts if the type is ranked but doesn't have the dimensio
tao_compiler/mlir/disc/transforms/lower_tf.cc:886
↓ 6 callersFunctiongetDimSizesOfTensor
tao_compiler/mlir/disc/disc_util.h:37
↓ 6 callersFunctiongetNonSizeOneDims
tao_compiler/mlir/disc/transforms/fusion_utils.cc:2040
↓ 6 callersFunctiongetScalarTypeFromValue
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:107
↓ 6 callersFunctionget_canonical_input_name
(input_name: str)
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:595
↓ 6 callersFunctionget_node_by_name
( graph_def: Optional[tf.GraphDef], simple_graph: SimpleGraph, name: str )
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:633
↓ 6 callersFunctionget_padding_size
(length, dap_size)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/utils.py:21
↓ 6 callersFunctiongroup_nodes
(block)
pytorch_blade/torch_blade/clustering/support_group_conversion.py:191
↓ 6 callersMethodh2d
tao_compiler/mlir/xla/ral/device/gpu/gpu_driver.cc:174
↓ 6 callersMethodhas_cycle
(self)
pytorch_blade/torch_blade/clustering/support_fusion_algorithm.py:49
↓ 6 callersMethodhash
tao_compiler/mlir/disc/transforms/shape_utils.h:142
↓ 6 callersFunctioninsertTranspose
tao_compiler/mlir/disc/transforms/disc_transpose_simplifier.cc:260
↓ 6 callersFunctioninternal_root_dir
()
scripts/python/common_setup.py:534
← previousnext →301–400 of 6,088, ranked by callers