MCPcopy Create free account

hub / github.com/alibaba/BladeDISC / functions

Functions6,088 in github.com/alibaba/BladeDISC

⨍ Functions 6,088◇ Types & classes 1,279↳ Endpoints 6
↓ 2 callersFunctionjson_num_add
tao/tao_bridge/kernels/tao_compilation_info_collector.cc:36
↓ 2 callersFunctionlang_type
(filename)
scripts/pre-commit/copyright.py:68
↓ 2 callersMethodlexical_order_topolist
(self)
pytorch_blade/torch_blade/algorithm/directed_graph.py:92
↓ 2 callersFunctionlower
()
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/utils.py:197
↓ 2 callersFunctionlowerGatherOpInternal
tao_compiler/mlir/disc/transforms/lhlo_elemental_utils.cc:700
↓ 2 callersFunctionlowerWithScheduleLargeConcatCPU
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:4149
↓ 2 callersFunctionlowerWithScheduleLoopCPU
we don't do inbound check for kLoop Schedule LoopSplit pass will do this. * loop.for %idxs... in shape_of(dominant_op) { * %linear_idx = getLinear
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:5207
↓ 2 callersFunctionlowerWithScheduleSparseSegmentReductionWithEmptyRowsOpCPU
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:4861
↓ 2 callersFunctionlowerWithScheduleWhereOpCPU
tao_compiler/mlir/disc/transforms/lhlo_legalize_roots_to_loops.cc:5076
↓ 2 callersFunctionmakeConvTuningCacheKey
tao_compiler/mlir/xla/ral/context/stream_executor_based_impl.cc:819
↓ 2 callersFunctionmakeNewConvParams
tao_compiler/mlir/xla/ral/context/stream_executor_based_impl.cc:1198
↓ 2 callersFunctionmakeTransposeOp
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:1064
↓ 2 callersMethodmake_fusion_group_executable
This function is used to make the fusion group executable. Before this function, the graph is like: graph(%se
pytorch_blade/torch_blade/quantization/prepare_data.py:73
↓ 2 callersMethodmake_region
tao_compiler/mlir/xla/ral/context/common_context_impl_acl.cc:117
↓ 2 callersFunctionmapStrToType
pytorch_blade/tests/torch-disc-pdll/utils.cpp:446
↓ 2 callersMethodmapSymbolicDimProductEqual
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:313
↓ 2 callersMethodmarkOwned
mark that this wrapper exclusively owns the underlying buffer.
tao_compiler/mlir/xla/ral/context/base/base_context.h:60
↓ 2 callersFunctionmaybeCastTo
Returns a new scalar integer value having type `type`. Here `type` must be an integer or index type.
tao_compiler/mlir/disc/IR/hlo_disc_ops.cc:456
↓ 2 callersFunctionmergeTypes
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:128
↓ 2 callersMethodmerge_node
(self, u: int, v: int)
pytorch_blade/torch_blade/algorithm/directed_graph.py:29
↓ 2 callersMethodmnit
(self, device)
pytorch_blade/tests/ltc/test_mnist.py:94
↓ 2 callersFunctionmodel_training_evaluation
( backend, train_dataloader, eval_dataloader, model, optimizer, num_epochs, evaluation)
examples/PyTorch/Train/Lazy/Bert/bert_train_ltc.py:123
↓ 2 callersFunctionmodel_training_evaluation
( backend, train_dataloader, eval_dataloader, model, optimizer, num_epochs, evaluation, is_profiling=False
examples/PyTorch/Train/Dynamo/Bert/test_bert.py:125
↓ 2 callersMethodnewSymbolicDim
tao_compiler/mlir/disc/transforms/disc_shape_optimization_utils.cc:100
↓ 2 callersMethodnoGuardCondition
tao_compiler/mlir/disc/transforms/disc_transform_schedule.cc:1452
↓ 2 callersMethodnum_allocations
tao/tao_bridge/executable.h:87
↓ 2 callersFunctionnum_make_jobs
()
scripts/python/common_setup.py:357
↓ 2 callersFunctionnum_make_jobs
()
pytorch_blade/common_setup.py:357
↓ 2 callersFunctionoffsetAllZeros
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:232
↓ 2 callersFunctionoss_eps
()
scripts/python/bazel_snapshot_client.py:72
↓ 2 callersFunctionparseEntryFunctionInputPlacements
tao_compiler/mlir/disc/transforms/placement_utils.cc:203
↓ 2 callersFunctionparseEntryFunctionOutputPlacements
tao_compiler/mlir/disc/transforms/placement_utils.cc:209
↓ 2 callersFunctionparseInnerMostDimsInfo
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:1034
↓ 2 callersFunctionparsePlacementAttribute
tao_compiler/mlir/disc/transforms/placement_utils.cc:155
↓ 2 callersFunctionparse_permute_dims
pytorch_blade/pytorch_blade/compiler/jit/torch/eliminate_redundant_permutations.cpp:28
↓ 2 callersFunctionparsefusionTagSetFromStr
tao_compiler/mlir/disc/transforms/fusion_utils.cc:288
↓ 2 callersFunctionpartitionRunner
tao_compiler/mlir/xla/ral/context/common_context_impl.cc:913
↓ 2 callersFunctionpatch_chunk_func
(model, func_name)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/optimization/torchscript_optimize.py:65
↓ 2 callersFunctionpatch_func
(old_func)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/optimization/torchscript_optimize.py:46
↓ 2 callersFunctionpatch_model_forward
patch forward func to run tasks
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/optimization/utils.py:21
↓ 2 callersFunctionprintStats
(backend, timings, batch_size=1)
examples/PyTorch/Inference/CPU/albert/main.py:64
↓ 2 callersFunctionpropagateNode
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:629
↓ 2 callersFunctionpropagateTorchTensorShape
pytorch_blade/pytorch_blade/compiler/jit/torch/shape_analysis.cpp:601
↓ 2 callersMethodpy_run
(self)
pytorch_blade/setup.py:87
↓ 2 callersFunctionral_base_cuda_const_cuda_internal
tao_compiler/mlir/xla/ral/context/common_context_impl_cuda.cc:59
↓ 2 callersFunctionral_base_cuda_const_host_internal
tao_compiler/mlir/xla/ral/context/common_context_impl.cc:279
↓ 2 callersFunctionral_tf_send_output_impl
tao_compiler/mlir/xla/ral/context/tensorflow/tf_context_impl.cc:759
↓ 2 callersFunctionral_to_bool
tao_compiler/mlir/xla/ral/context/tensorflow/tf_context_impl.h:34
↓ 2 callersMethodrankedTensor2SymDims
tao_compiler/mlir/disc/transforms/disc_shape_optimization.cc:1452
↓ 2 callersFunctionrecord_shape_by_tracing
(module, inputs, graph=None)
pytorch_blade/torch_blade/tools/shape_inference.py:122
↓ 2 callersFunctionreduce_edge_list
(l)
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/utils.py:146
↓ 2 callersFunctionregisterAllDiscPasses
tao_compiler/mlir/disc/transforms/register_passes.h:28
↓ 2 callersFunctionregisterAllMhloDiscPasses
tao_compiler/mlir/disc/transforms/register_passes.h:39
↓ 2 callersFunctionregisterPredefinedHelperFunctions
tao_compiler/mlir/disc/transforms/disc_pdl_utils.cc:370
↓ 2 callersFunctionregisterTilingInterfaceExternalModels
tao_compiler/mlir/disc/tools/disc-transform/LinalgExt/LinalgExtOps.cc:875
↓ 2 callersFunctionregisterTransformDialectCommonExtension
tao_compiler/mlir/disc/tools/disc-transform/TransformOps/TransformOpsExt.cc:3231
↓ 2 callersFunctionregister_forward_hook
register forward func
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/optimization/utils.py:41
↓ 2 callersMethodreifyReturnTypeShapes
tao_compiler/mlir/disc/IR/hlo_disc_ops.cc:74
↓ 2 callersMethodreleaseAllFreeBuffers
tao_compiler/mlir/xla/ral/context/base/base_context.cc:102
↓ 2 callersMethodreleaseDeviceConstant
tao_compiler/mlir/xla/ral/ral_metadata.cc:116
↓ 2 callersFunctionremove_node_by_index
( graph_def: tf.GraphDef, remove_list: Union[Set[int], List[int]] )
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:708
↓ 2 callersFunctionrename_node_inputs
(graph_def: tf.GraphDef, rename_dict: Dict[str, str])
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:719
↓ 2 callersFunctionreplace_group_with_engine
( graph, module_holder, node, attr_name, eng_type, group_input
pytorch_blade/torch_blade/clustering/support_group_conversion.py:21
↓ 2 callersMethodresource_var_operation_node_ids
Returns the TF graph node IDs for the resource variable operations in this cluster.
tao/tao_bridge/passes/tao_mark_for_compilation_pass.cc:221
↓ 2 callersFunctionrestore_args_from_cache
Restore confs from `disc_conf_file()`
scripts/python/common_setup.py:734
↓ 2 callersFunctionrestore_args_from_cache
Restore confs from `disc_conf_file()`
pytorch_blade/common_setup.py:734
↓ 2 callersFunctionrewriteValue
tao_compiler/mlir/disc/transforms/disc_custom_call_rewriter.cc:53
↓ 2 callersMethodrun
tao_compiler/mlir/disc/transforms/disc_dot_merge.cc:216
↓ 2 callersMethodrun
pytorch_blade/pytorch_blade/compiler/jit/torch/freeze_module.cpp:125
↓ 2 callersFunctionrun_benchmark
(graph_def, fetches, feed_dict, model_name)
tensorflow_blade/examples/Inference/CUDA/BERT/bert_inference_opt.py:58
↓ 2 callersFunctionrun_tensorrt
(model, input_tensors, params, precision, is_trt_engine=False)
pytorch_blade/benchmark/torch-tensorrt/perf_run.py:199
↓ 2 callersFunctionrun_torch
(model, input_tensors, params, precision)
pytorch_blade/benchmark/torch-tensorrt/perf_run.py:71
↓ 2 callersFunctionrun_torch_blade
(model, input_tensors, params, precision)
pytorch_blade/benchmark/torch-tensorrt/perf_run.py:100
↓ 2 callersFunctionrun_torch_tensorrt
(model, input_tensors, params, precision)
pytorch_blade/benchmark/torch-tensorrt/perf_run.py:137
↓ 2 callersFunctionsafe_run
(cmd, shell=False, verbose=True)
pytorch_blade/common_setup.py:512
↓ 2 callersMethodsame_group
(self, src: int, dst: int)
pytorch_blade/torch_blade/clustering/support_fusion_algorithm.py:28
↓ 2 callersFunctionsetFusionName
Sets the name of the fusion op
tao_compiler/mlir/disc/transforms/fusion_utils.cc:254
↓ 2 callersMethodsetSymbolDim
tao_compiler/mlir/disc/transforms/shape_utils.h:67
↓ 2 callersMethodsetUp
(self)
pytorch_blade/tests/disc/pdl/test_e2e/test_quantization.py:48
↓ 2 callersFunctionset_attr_float
(node: tf.NodeDef, key: str, value: float)
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:120
↓ 2 callersFunctionset_attr_shape
( node: tf.NodeDef, key: str, value: Union[List[int], Tuple[int]] )
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:142
↓ 2 callersMethodset_options
tao/tao_bridge/passes/tao_optimization_pass.h:52
↓ 2 callersFunctionset_qconfig
(ctx: GraphModContext)
tools/torch_quant/torch_quant/graph.py:206
↓ 2 callersFunctionset_tensor_shape
pytorch_blade/pytorch_blade/compiler/jit/fusion.cpp:139
↓ 2 callersFunctionsetattr_leaf
(obj, fields, val)
pytorch_blade/torch_blade/tensorrt/dynamic_shapes.py:132
↓ 2 callersMethodsize
tao_compiler/mlir/xla/ral/context/custom_library/philox_random.h:60
↓ 2 callersFunctionsoftmax
Softmax, but without automatic casting to fp32 when the input is of type bfloat16
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/ops.py:257
↓ 2 callersMethodsort
in-place HEAPSORT in Cormen This method destroys the heap property.
tao_compiler/mlir/xla/ral/context/custom_library/tf_topk.cu.h:172
↓ 2 callersFunctionsymlink_disc_files
(args)
pytorch_blade/common_setup.py:606
↓ 2 callersFunctiontao_ci_conf_file
()
scripts/python/tao_build.py:112
↓ 2 callersMethodtarget_device
tao/tao_bridge/mlir/mlir_executable.cc:76
↓ 2 callersFunctiontensor_name_to_node_name
(tensor_name: str)
tensorflow_blade/tf_blade/util/tf_util.py:81
↓ 2 callersFunctiontest_tao_bridge
(root, args, cpp=True, python=True)
scripts/python/tao_build.py:522
↓ 2 callersFunctiontest_tao_compiler_add_flags_platform_alibaba
(root, args, flag)
scripts/python/common_setup.py:600
↓ 2 callersFunctiontileParallelLoop
Tile a parallel loop of the form scf.parallel (%i0, %i1) = (%arg0, %arg1) to (%arg2, %arg3) step (%arg4, %arg5) into scf.parallel (%i0, %i1) = (%arg0
tao_compiler/mlir/disc/transforms/codegen_utils.cc:204
↓ 2 callersFunctiontoJSON
tao_compiler/mlir/disc/IR/topk_custom_call_op.cc:39
↓ 2 callersFunctiontransform_graph
( graph_def: tf.GraphDef, inputs: List[str], outputs: List[str], transforms: List[str] )
tensorflow_blade/tf_blade/util/tf_graph_transform_util.py:960
↓ 2 callersFunctiontrim
trim from both ends (in place)
pytorch_blade/tests/torchscript/shape_analysis_tool.cpp:43
↓ 2 callersFunctiontrunc_normal_init_
(weights, scale=1.0, fan="fan_in")
examples/PyTorch/Inference/CUDA/AlphaFold/FoldAcc/foldacc/model/modules/ops.py:46
↓ 2 callersFunctiontryMoveUpTransposeGreedilyAndApplyIfBeneficialImpl
tao_compiler/mlir/disc/transforms/disc_transpose_simplifier.cc:758
← previousnext →1,201–1,300 of 6,088, ranked by callers