MCPcopy Create free account

hub / github.com/alibaba/BladeDISC / functions

Functions6,088 in github.com/alibaba/BladeDISC

↓ 1 callersMethodisFullySlicedAxis
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:927
↓ 1 callersMethodisFusableDefault
Default fusability check - used when the user doesn't pass in a callback.
pytorch_blade/pytorch_blade/ltc/disc_compiler/passes/graph_fuser.cpp:209
↓ 1 callersMethodisFusible
tao_compiler/mlir/disc/transforms/fusion_utils_sparse_op_cpu.cc:42
↓ 1 callersMethodisInteriorPaddingUnknown
Returns true if the `interiorPadding` of the `axis` is unknown.
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.h:226
↓ 1 callersFunctionisMutableTypeImpl
pytorch_blade/pytorch_blade/compiler/jit/torch/alias_analysis.cpp:64
↓ 1 callersMethodisNotPaddedAxis
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:1035
↓ 1 callersFunctionisOpTriviallyDeadDisc
pytorch_blade/pytorch_blade/torch-mlir/lib/Dialect/TorchConversion/Transforms/ApplyDiscPdlPatterns.cpp:100
↓ 1 callersFunctionisQint32ConstantZeroTensor
tao_compiler/mlir/disc/transforms/lower_tf.cc:758
↓ 1 callersMethodisShapeValueEqual
tao_compiler/mlir/disc/transforms/shape_utils.cc:849
↓ 1 callersFunctionisSimpleMap
What is a simple mappable operator? It: - Has a single tensor output - Output and all tensor inputs have the same shape - Output and all tensor input
pytorch_blade/pytorch_blade/ltc/disc_compiler/passes/graph_fuser.cpp:46
↓ 1 callersFunctionisSingleElementH2DOp
tao_compiler/mlir/disc/transforms/fusion_utils.cc:1445
↓ 1 callersFunctionisSinkingBeneficiary
Identifies operations that are beneficial to sink into kernels. These operations may not have side-effects, as otherwise sinking (and hence duplicatin
tao_compiler/mlir/disc/transforms/disc_outline_cpu_kernel.cc:48
↓ 1 callersFunctionisSmallCpuKernel
tao_compiler/mlir/disc/transforms/disc_cpu_map_parallel_loop.cc:81
↓ 1 callersFunctionisSmallForOp
tao_compiler/mlir/disc/transforms/disc_cpu_map_parallel_loop.cc:65
↓ 1 callersMethodisStartIndexUnknown
Returns true if the `startIndices` of the `axis` is unknown.
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.h:134
↓ 1 callersFunctionisStitchEnabled
tao_compiler/mlir/disc/disc_util.cc:263
↓ 1 callersMethodisStrideUnknown
Returns true if the `stride` of the `axis` is unknown.
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.h:155
↓ 1 callersMethodisSymbolicDimProductEqual
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:300
↓ 1 callersFunctionisTensorDialect
Check Op if it is a tensor Op.
tao_compiler/mlir/disc/transforms/placement_utils.h:92
↓ 1 callersMethodisUnfold
tao_compiler/mlir/disc/transforms/shape_utils.h:110
↓ 1 callersMethodisWrittenBack
tao_compiler/mlir/disc/transforms/lhlo_elemental_utils.h:122
↓ 1 callersMethodis_4c_blocked
workaround for issue intel/mkl-dnn#588
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:195
↓ 1 callersFunctionis_aarch64
()
scripts/python/common_setup.py:353
↓ 1 callersFunctionis_available
()
pytorch_blade/torch_blade/mlir/__init__.py:36
↓ 1 callersFunctionis_available
()
pytorch_blade/torch_blade/neural_engine/__init__.py:25
↓ 1 callersFunctionis_concrete_shape_tensor_type
pytorch_blade/pytorch_blade/compiler/jit/tool_funcs.cpp:88
↓ 1 callersMethodis_default
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:142
↓ 1 callersMethodis_entry_computation_parameter
tao/tao_bridge/executable.h:92
↓ 1 callersMethodis_grouped
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:424
↓ 1 callersMethodis_hwio
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:184
↓ 1 callersMethodis_maybe_live_out
tao/tao_bridge/executable.h:97
↓ 1 callersFunctionis_per_channel
(qscheme: torch.qscheme)
tools/torch_quant/torch_quant/observer.py:94
↓ 1 callersFunctionis_per_tensor
(qscheme: torch.qscheme)
tools/torch_quant/torch_quant/observer.py:90
↓ 1 callersMethodis_plain
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:138
↓ 1 callersMethodis_resource
tao/tao_bridge/executable.h:59
↓ 1 callersFunctionis_symmetric
(qscheme: torch.qscheme)
tools/torch_quant/torch_quant/observer.py:98
↓ 1 callersMethodis_temp_buffer
tao/tao_bridge/executable.h:112
↓ 1 callersMethodis_thread_local
tao/tao_bridge/executable.h:116
↓ 1 callersMethodis_tuple
tao/tao_bridge/executable.h:107
↓ 1 callersFunctionis_x86
()
scripts/python/common_setup.py:348
↓ 1 callersFunctionis_x86
()
pytorch_blade/common_setup.py:348
↓ 1 callersMethodis_zero
returns true if memory descriptor is zero */
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/tensor.hpp:127
↓ 1 callersFunctioniskSparseReductionOutputFusible
tao_compiler/mlir/disc/transforms/fusion_utils_sparse_op_cpu.cc:30
↓ 1 callersFunctioniskWhereInputFusible
tao_compiler/mlir/disc/transforms/fusion_utils_sparse_op_cpu.cc:25
↓ 1 callersFunctionjson_obj_add
tao/tao_bridge/kernels/tao_compilation_info_collector.cc:44
↓ 1 callersFunctionlayout_match
tao_compiler/mlir/xla/ral/context/stream_executor_based_impl.cc:1570
↓ 1 callersFunctionlecun_normal_init_
(weights)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/ops.py:60
↓ 1 callersFunctionlink_files
(args)
tensorflow_blade/build.py:113
↓ 1 callersFunctionlist_files_with_extensions
(dir, extensions)
examples/PyTorch/Inference/CUDA/AlphaFold/openfold_inference.py:59
↓ 1 callersFunctionloadArguments
pytorch_blade/pytorch_blade/ltc/disc_compiler/replay.cpp:64
↓ 1 callersFunctionloadDependentDialects
tao_compiler/mlir/disc/tools/disc-source-emitter/disc-cuda-emitter.cc:25
↓ 1 callersFunctionloadDependentDialects
tao_compiler/mlir/disc/tools/disc-pdll/disc-pdll.cc:39
↓ 1 callersFunctionloadDependentDialects
pytorch_blade/tests/torch-disc-pdll/torch-disc-pdll.cpp:29
↓ 1 callersFunctionloadGlobalString
Loads a global op at the current insertion point and returns the loaded value.
tao_compiler/mlir/disc/transforms/disc_to_llvm.cc:158
↓ 1 callersFunctionloadGraph
pytorch_blade/pytorch_blade/ltc/disc_compiler/replay.cpp:56
↓ 1 callersFunctionload_feature_for_one_target
( config, data_folder, seed=0, is_multimer=False, use_uniprot=False )
examples/PyTorch/Inference/CUDA/AlphaFold/unifold_inference.py:48
↓ 1 callersFunctionload_frozen_graph
(model_file : str)
examples/TensorFlow/Inference/X86/BERT/main.py:21
↓ 1 callersFunctionload_frozen_graph
(model_file : str)
examples/TensorFlow/Inference/AArch64/BERT/main.py:21
↓ 1 callersFunctionload_frozen_graph
(model_file: str)
examples/TensorFlow/Inference/CUDA/BERT/main.py:23
↓ 1 callersFunctionload_model
()
examples/PyTorch/Inference/CPU/albert/main.py:151
↓ 1 callersFunctionload_model
(params)
pytorch_blade/benchmark/torch-tensorrt/perf_run.py:369
↓ 1 callersFunctionlogger_level_context
(level)
pytorch_blade/torch_blade/logging.py:24
↓ 1 callersFunctionloopUnrollByFactorAndTryInterleave
This function unrolls the given for-loop op and interleaves the instructions
tao_compiler/mlir/disc/transforms/codegen_utils.cc:524
↓ 1 callersFunctionlowerFakeQuantToQuantAndDequant
tao_compiler/mlir/disc/disc_util.cc:243
↓ 1 callersFunctionlowerWithScheduleLoopV2
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:332
↓ 1 callersFunctionlowerWithScheduleSparseFillEmptyRowsOpCPU
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:4375
↓ 1 callersFunctionlowerWithScheduleSparseReshapeOpCPU
SparseReshape implementation: Loop over input for (i = 0; i < input_indices.shape(0), ++i) { // multidim index to linear index linear_index = ...; /
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:4201
↓ 1 callersFunctionlowerWithScheduleSparseSegmentReductionOpCPU
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:4631
↓ 1 callersFunctionlowerWithScheduleStitch
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:2883
↓ 1 callersFunctionlowerWithScheduleStitchV2
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:3564
↓ 1 callersFunctionlowerWithSchedulekInputCPU
Emitter for non-row-reduction kInput fusion pattern. Take a column reduction `memref<100x1100xf32> -> memref<1100xf32>` as an example: for (int iOutt
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:5294
↓ 1 callersFunctionmain
()
tao_compiler/mlir/xla/ral/test/bace.py:25
↓ 1 callersFunctionmain
()
tao_compiler/mlir/xla/ral/test/codgen_kernels.py:25
↓ 1 callersFunctionmain
tao_compiler/mlir/disc/disc_compiler_main.cc:241
↓ 1 callersFunctionmain
()
scripts/python/tao_build.py:858
↓ 1 callersFunctionmain
()
scripts/python/bazel_snapshot_client.py:144
↓ 1 callersFunctionmain
(argv=None)
scripts/pre-commit/copyright.py:88
↓ 1 callersFunctionmain
()
examples/PyTorch/Train/Lazy/Bert/bert_train_ltc.py:227
↓ 1 callersFunctionmain
()
examples/PyTorch/Train/Dynamo/Bert/test_bert.py:255
↓ 1 callersFunctionmain
(args)
examples/PyTorch/Inference/CUDA/AlphaFold/openfold_inference.py:62
↓ 1 callersFunctionmain
(args)
examples/PyTorch/Inference/CUDA/AlphaFold/unifold_inference.py:75
↓ 1 callersFunctionmain
()
tensorflow_blade/build.py:385
↓ 1 callersFunctionmakeDynamicShapeGEMMWithBiasParamsKey
tao_compiler/mlir/xla/ral/context/common_context_impl_mkldnn.h:235
↓ 1 callersFunctionmakeGEMMWithBiasParamsKey
tao_compiler/mlir/xla/ral/context/common_context_impl_mkldnn.h:200
↓ 1 callersFunctionmakeGemmTuningCacheKey
tao_compiler/mlir/xla/ral/context/stream_executor_based_impl.cc:92
↓ 1 callersFunctionmakeNewDeviceStrategy
tao_compiler/mlir/disc/transforms/fusion_utils.cc:1561
↓ 1 callersFunctionmakeSortDescriptor
tao_compiler/mlir/xla/ral/context/dynamic_sort_impl.h:26
↓ 1 callersFunctionmakeSymbolicDimOpRefArrayAttr
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:779
↓ 1 callersFunctionmake_filename
This function is copied from c10/util/tempfile.h, so it follows to these temperary directory env variables, too.
pytorch_blade/pytorch_blade/common_utils/tempfs.cpp:28
↓ 1 callersFunctionmake_package
(root, args)
scripts/python/tao_build.py:632
↓ 1 callersMethodmalloc
tao_compiler/mlir/xla/ral/context/mkldnn/ideep/ideep/allocators.hpp:13
↓ 1 callersFunctionmapResultToInitOperandOfDestinationStyleOp
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:271
↓ 1 callersMethodmarkAsFullySlicedAxis
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:933
↓ 1 callersFunctionmatchAndRwriterQuantizedConv2DWithBiasAndRequantizeOp
tao_compiler/mlir/disc/transforms/lower_tf.cc:1033
↓ 1 callersFunctionmaterializeShapeComputation
materialize shape computation IR by resolving tensor.dim + op having interface InferShapedTypeOpInterface: before ``` %arg0 = ... : tensor<?x10xf32> %
tao_compiler/mlir/disc/transforms/disc_shape_optimization.cc:235
↓ 1 callersFunctionmayConvertToIntegerType
tao_compiler/mlir/disc/transforms/codegen_utils.cc:148
↓ 1 callersFunctionmaybeEmitInitLoops
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:870
↓ 1 callersMethodmerge
Merges `other` into this cluster, and clears `other`.
tao_compiler/mlir/disc/transforms/disc_quantized_dot_merge.cc:66
↓ 1 callersMethodmerge
Merges `other` into this cluster, and clears `other`.
tao_compiler/mlir/disc/transforms/disc_dot_merge.cc:84
↓ 1 callersFunctionmergeAssignment
tao_compiler/mlir/disc/transforms/disc_assign_memory_space.cc:73
← previousnext →2,301–2,400 of 6,088, ranked by callers