MCPcopy Create free account

hub / github.com/alibaba/BladeDISC / functions

Functions6,088 in github.com/alibaba/BladeDISC

↓ 3 callersMethodregister_pipeline
(cls, name, func)
pytorch_blade/torch_blade/config.py:28
↓ 3 callersMethodreleaseHostConstant
tao_compiler/mlir/xla/ral/ral_metadata.cc:101
↓ 3 callersMethodreplace_root
tao_compiler/mlir/xla/ral/context/custom_library/tf_topk.cu.h:181
↓ 3 callersFunctionrestore_env_var
(name, saved)
scripts/python/tao_common.py:128
↓ 3 callersFunctionrunning_on_ci
Return true if the building job is running on CI host.
scripts/python/common_setup.py:198
↓ 3 callersFunctionrunning_on_ci
Return true if the building job is running on CI host.
pytorch_blade/common_setup.py:198
↓ 3 callersMethodsetSpecificLoader
tao_compiler/mlir/disc/transforms/lhlo_elemental_utils.h:126
↓ 3 callersMethodsetUpWithMLIR
(testcase='')
tao/tao_bridge/test/gpu/mlir/base.py:27
↓ 3 callersMethodset_allocation
tao/tao_bridge/executable.h:125
↓ 3 callersMethodset_deleter
tao_compiler/mlir/xla/ral/context/base/base_context.h:54
↓ 3 callersFunctionset_extra_inputs
(c_module, extra_inputs, extra_dynamic_shapes)
pytorch_blade/torch_blade/tensorrt/dynamic_shapes.py:131
↓ 3 callersMethodset_scale
Set new scale into param
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:895
↓ 3 callersMethodset_zero_point
Set new scale into param
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:911
↓ 3 callersFunctionstr2format
tao_compiler/mlir/xla/ral/context/common_context_impl_mkldnn.cc:162
↓ 3 callersFunctionsymlink_disc_files
(args)
scripts/python/common_setup.py:606
↓ 3 callersFunctiontao_bridge_bazel_config
(args)
scripts/python/tao_build.py:451
↓ 3 callersFunctiontao_bridge_dir
(root=None)
scripts/python/common_setup.py:537
↓ 3 callersFunctiontoPlacementType
tao_compiler/mlir/disc/transforms/mhlo_placer.cc:38
↓ 3 callersMethodtoString
pytorch_blade/pytorch_blade/ltc/disc_backend/backend_impl.cpp:56
↓ 3 callersMethodto_functiondef
Generate subgraph FunctionDef from this segment. Parameters ---------- seg_id : int = 0 The id of this segment/su
tensorflow_blade/tf_blade/util/simple_graph.py:459
↓ 3 callersMethodtopological_sort
Sort given SimpleGraph in topological order. Parameters ---------- reverse : bool = False Set True to list op fro
tensorflow_blade/tf_blade/util/simple_graph.py:167
↓ 3 callersFunctiontrace_model
(model: PreTrainedModel, info: ModelInfo, amp: bool = False)
examples/PyTorch/Inference/hf_transformers/blade_adapter.py:256
↓ 3 callersMethodupdate
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/lru_cache.hpp:155
↓ 3 callersFunctionupdateFunctionType
Updates the function types according to the types of entry block arguments and the types of operands of the return op of the func op. This function su
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:800
↓ 3 callersFunctionwarpFilterCompress
tao_compiler/mlir/xla/ral/context/custom_library/warp.cu.h:144
↓ 3 callersFunctionwarpReduceMaxWithPos
tao_compiler/mlir/xla/ral/context/custom_library/warp.cu.h:62
↓ 3 callersFunctionwrap_model_inputs
wrap forward func to ignore some inputs
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/optimization/utils.py:93
↓ 3 callersMethodwrite_version_file
(self, version_path)
pytorch_blade/torch_blade_build.py:112
↓ 2 callersFunctionAddDtypeToKernalDefConstraint
tao/tao_bridge/tf/xla_op_registry.cc:628
↓ 2 callersMethodAddExecutable
tao/tao_bridge/kernels/tao_compilation_info_collector.cc:670
↓ 2 callersFunctionAddMissingFunctionByName
Copy the FunctionDef of given function from lookup_library to library, if it can be found in lookup_library but is missing from library.
tao/tao_bridge/passes/functionalize_while.cc:237
↓ 2 callersFunctionAddMissingFunctionDef
Iterate over all functions that the given fdef refers to. Copy the missing FunctionDefs from lookup_library to library.
tao/tao_bridge/passes/functionalize_while.cc:248
↓ 2 callersFunctionAlign
std::align is not supported, so this method mimic its behavior.
tensorflow_blade/src/util/tf_allocator_util.cc:87
↓ 2 callersMethodAllNodesInPostOrder
tao/tao_bridge/tf/graphcycles.cc:432
↓ 2 callersMethodAllocateDeviceMemory
tensorflow_blade/src/tensorrt/bridge/tensorrt_tf_allocator.cc:88
↓ 2 callersMethodAncestorStateToString
tao/tao_bridge/passes/functionalize_cond.cc:235
↓ 2 callersFunctionAppendMarkForCompilationPassFlagsInternal
tao/tao_bridge/tf/flags.cc:76
↓ 2 callersFunctionArrangeOperandsInsertPointInBlock
Arrange the insert-point of `op`'s operands in the same block. Do not deal with cross-block operands.
tao_compiler/mlir/disc/transforms/disc_dot_merge.cc:42
↓ 2 callersFunctionBroadcastTensorRanks
pytorch_blade/pytorch_blade/torch-mlir/lib/Conversion/TorchToMhlo/DiscTorchToMhlo.cpp:56
↓ 2 callersFunctionBuildArgNode
tao/tao_bridge/passes/functionalize_while.cc:138
↓ 2 callersMethodBuildEngine
tensorflow_blade/src/tensorrt/bridge/tensorrt_onnx_parser.cpp:72
↓ 2 callersFunctionCallStatusReturningFunction
tao/tao_bridge/errors_test.cc:71
↓ 2 callersMethodCanContractEdge
tao_compiler/mlir/disc/utils/cycle_detector.cc:346
↓ 2 callersFunctionCandidateCudaRoots
tao_compiler/mlir/disc/transforms/legalize_to_cubin.cc:91
↓ 2 callersFunctionCeilOfRatio
tao_compiler/mlir/xla/ral/context/custom_library/transpose_gpu.cu.cc:21
↓ 2 callersFunctionCheckEdges
tao/tao_bridge/tf/graphcycles_test.cc:158
↓ 2 callersFunctionCheckNodeNotInCycle
Check that the graph has no cycle containing the given node.
tao/tao_bridge/passes/functionalize_control_flow_util.cc:58
↓ 2 callersFunctionCheckTransitiveClosure
tao/tao_bridge/tf/graphcycles_test.cc:135
↓ 2 callersMethodClear
tao/tao_bridge/tf/ordered_set.h:68
↓ 2 callersFunctionClearVisitedBits
tao/tao_bridge/tf/graphcycles.cc:291
↓ 2 callersFunctionClearVisitedBits
Clears bookkeeping fileds used during the last DFS process.
tao_compiler/mlir/disc/utils/cycle_detector.cc:248
↓ 2 callersFunctionCommonVerifyForQuantizedComputeIntensiveOp
tao_compiler/mlir/disc/IR/hlo_disc_ops.cc:352
↓ 2 callersMethodCompile
tao_compiler/mlir/disc/tools/disc-replay/disc_interpreter.cc:26
↓ 2 callersFunctionCompileToDiscExecutable
pytorch_blade/pytorch_blade/ltc/disc_compiler/disc_compiler.cpp:121
↓ 2 callersFunctionCondConstInputIndices
tao/tao_bridge/tf/const_analysis.cc:62
↓ 2 callersFunctionConvertResults
Converts the results of the operation `op` to memref types and append them to the `results` vector.
tao_compiler/mlir/disc/transforms/disc_hlo_legalize_to_lhlo.cc:105
↓ 2 callersFunctionConvertTF2MlirHlo
tao_compiler/mlir/disc/disc_compiler.cc:941
↓ 2 callersFunctionCopySubgraph
Copies a subgraph from `graph` to `output` by performing a reverse DFS starting at nodes in vector `stack`. `node_map` is a vector indexed by source n
tao/tao_bridge/passes/functionalize_while.cc:100
↓ 2 callersFunctionCustomFuseGraph
pytorch_blade/pytorch_blade/ltc/disc_compiler/passes/graph_fuser.cpp:1302
↓ 2 callersFunctionDataTypeToMlirType
tao_compiler/decoupling/mlir_compiler.cc:151
↓ 2 callersFunctionDescribeCycle
Returns a string describing how an edge from src to dst would create a cycle.
tao/tao_bridge/tf/xla_cluster_util.cc:53
↓ 2 callersFunctionDeviceTypeToStr
tao_compiler/mlir/disc/tests/mlir_feature_test.cc:46
↓ 2 callersFunctionDoAsserts
tao/tao_bridge/dumper_common_test.cc:47
↓ 2 callersFunctionDoAsserts
tao/tao_bridge/common_test.cc:47
↓ 2 callersFunctionDumpGraphDefToFile
tao/tao_bridge/tf/dump_graph.cc:91
↓ 2 callersFunctionDumpLLVMModule
tao_compiler/mlir/disc/disc_compiler.cc:84
↓ 2 callersMethodEmitOp
tao_compiler/mlir/disc/utils/source_emitter.cc:587
↓ 2 callersFunctionFindFirstNotOf
Like s.find_first_not_of(x, pos), but return s.size() when find_first_not_of() would return string::npos. This avoids if-statements elsewhere.
tao/tao_bridge/tf/parse_flags_from_env.cc:88
↓ 2 callersFunctionFindNodesToDecluster
tao/tao_bridge/passes/tao_partially_decluster_pass.cc:41
↓ 2 callersFunctionFindUniqueBackedge
tao/tao_bridge/tf/deadness_analysis.cc:1012
↓ 2 callersMethodForEach
tao/tao_bridge/tf/device_util.h:74
↓ 2 callersFunctionForwardDFS
tao/tao_bridge/tf/graphcycles.cc:200
↓ 2 callersFunctionForwardDFS
Follows the edges from producer to consumer and searchs if the node having rank `n` can reach the node having rank `upper_bound` using a DFS search. W
tao_compiler/mlir/disc/utils/cycle_detector.cc:147
↓ 2 callersFunctionFusionOutputs
pytorch_blade/pytorch_blade/ltc/disc_compiler/replay.cpp:43
↓ 2 callersFunctionGenerateMlirModuleString
pytorch_blade/pytorch_blade/compiler/mlir/converters/mhlo_conversion.cpp:54
↓ 2 callersFunctionGenerateStrSeq
tao_compiler/mlir/disc/tests/mlir_feature_test.cc:36
↓ 2 callersMethodGetBackendDevice
pytorch_blade/pytorch_blade/ltc/disc_backend/backend_impl.cpp:268
↓ 2 callersFunctionGetBackendName
pytorch_blade/pytorch_blade/compiler/tensorrt/tensorrt_engine.cpp:68
↓ 2 callersFunctionGetCandidateRowReduceOp
tao_compiler/mlir/disc/transforms/disc_specialize_fusion_with_speculation.cc:187
↓ 2 callersFunctionGetCheckPointId
tao/tao_bridge/passes/tao_mark_for_compilation_pass.cc:1656
↓ 2 callersFunctionGetCompileTimeConstInputs
tao/tao_bridge/tf/const_analysis.cc:100
↓ 2 callersFunctionGetCompilerFactories
tao_compiler/decoupling/compiler_base.cc:22
↓ 2 callersFunctionGetConvMetadata
tao_compiler/mlir/disc/transforms/disc_lower_to_library_call.cc:274
↓ 2 callersFunctionGetDebugNames
pytorch_blade/pytorch_blade/compiler/mlir/converters/mhlo_conversion.cpp:89
↓ 2 callersFunctionGetDiscCpuMathKernelMode
tao_compiler/mlir/xla/ral/context/common_context_impl_mkldnn.cc:125
↓ 2 callersMethodGetLeaderValueWithSameShapeInFusion
tao_compiler/mlir/disc/transforms/shape_utils.cc:1825
↓ 2 callersFunctionGetMicroBatchId
tao/tao_bridge/passes/tao_mark_for_compilation_pass.cc:1642
↓ 2 callersMethodGetOutsideCompilationSubgraphNames
tao/tao_bridge/passes/tao_encapsulate_subgraphs_pass.cc:1435
↓ 2 callersFunctionGetPointerToFuel
tao/tao_bridge/passes/tao_mark_for_compilation_pass.cc:2449
↓ 2 callersMethodGetRecvAtHostNode
tao/tao_bridge/passes/tao_encapsulate_subgraphs_pass.cc:731
↓ 2 callersFunctionGetRootFrame
If the node is inside some frames, get the name of the outermost non-empty frame. Otherwise, get an empty frame name.
tao/tao_bridge/tf/deadness_analysis.cc:1093
↓ 2 callersFunctionGetShapeFromConstUniqueName
tao_compiler/mlir/xla/ral/context/common_context_impl.cc:217
↓ 2 callersMethodGetShapeNum
tao/tao_bridge/kernels/tao_compilation_info_collector.h:110
↓ 2 callersFunctionGetSingletonState
tao/tao_bridge/tf/xla_config_registry.cc:32
↓ 2 callersMethodGetStatus
Returns the status. Used by MakeErrorStreamWithOutput.
tao/tao_bridge/errors.h:153
↓ 2 callersFunctionGetTensorAddress
tensorflow_blade/src/custom_ops/trt_engine_op/trt_engine_op.cc:57
↓ 2 callersFunctionGetXlaAutoClusteringSummary
tao/tao_bridge/tf/xla_cluster_util.cc:366
↓ 2 callersFunctionGpuLaunchKernel
tao_compiler/mlir/xla/ral/context/custom_library/gpu_helper.h:49
↓ 2 callersMethodHasOverflowed
tao/tao_bridge/tf/deadness_analysis.cc:603
↓ 2 callersFunctionHasResourceInputOrOutput
tao/tao_bridge/tf/xla_cluster_util.cc:215
← previousnext →801–900 of 6,088, ranked by callers