MCPcopy Create free account

hub / github.com/InfiniTensor/InfiniTensor / functions

Functions2,429 in github.com/InfiniTensor/InfiniTensor

↓ 4 callersMethodgetNumInputs
include/nnet/iterator_table.h:84
↓ 4 callersFunctiongetOffset
test/nnet/test_TConv2gemm.cc:278
↓ 4 callersMethodgetOps
include/core/graph_match.h:67
↓ 4 callersMethodgetReshapeDim
include/operators/transpose.h:36
↓ 4 callersMethodgetScales
include/operators/resize.h:134
↓ 4 callersMethodgetSourceRank
include/operators/send.h:30
↓ 4 callersMethodguidedDLT
src/nnet/Pass/Rule8GuidedDLT.cc:27
↓ 4 callersMethodhasData
include/core/tensor_base.h:37
↓ 4 callersMethodinit
src/operators/resize.cc:28
↓ 4 callersMethodinterpretUniformSample
src/nnet/Visitor/Interpreter.cc:153
↓ 4 callersMethodisMatMulOrConv
src/core/op_type.cc:280
↓ 4 callersMethodis_shard
(self, dim: Optional[int] = None)
examples/distributed/placement.py:8
↓ 4 callersMethodmatchWithPattern
src/nnet/Visitor/PatternMatcher.cc:116
↓ 4 callersMethodremoveTensor
include/core/graph.h:39
↓ 4 callersMethodreplaceSubGraph
replace all sub graphs which matched subA with subB in g
src/core/graph_match.cc:300
↓ 4 callersMethodrunWithCudaGraph
src/cuda/cuda_runtime.cc:42
↓ 4 callersMethodserialize
src/nnet/Visitor/Serializer.cc:76
↓ 4 callersFunctionserializeVec
include/nnet/expr.h:86
↓ 4 callersMethodsoftmax
src/core/graph_handler.cc:270
↓ 4 callersFunctiontest_layernormFp32
test/kernels/cuda/test_cuda_layernorm.cc:11
↓ 4 callersFunctiontest_reducemean
test/kernels/intelcpu/test_mkl_reduce.cc:11
↓ 4 callersFunctiontest_softmaxFp32
test/kernels/cuda/test_cuda_softmax.cc:11
↓ 4 callersFunctiontest_where
test/kernels/kunlun/test_kunlun_where.cc:10
↓ 4 callersFunctiontest_whereFp32
test/kernels/cuda/test_cuda_where.cc:11
↓ 4 callersMethodtoString
src/kunlun/kunlun_runtime.cc:58
↓ 4 callersMethodtune
(self)
pyinfinitensor/src/pyinfinitensor/onnx.py:1504
↓ 4 callersMethodtune
src/cuda/cuda_runtime.cc:63
↓ 3 callersMethodaddOp
src/core/graph_match.cc:13
↓ 3 callersMethodaddPredecessors
include/core/operator.h:124
↓ 3 callersMethodaddSuccessors
include/core/operator.h:125
↓ 3 callersFunctioncheck_exists
(path: str)
examples/distributed/kunlun/kunlun_launch.py:80
↓ 3 callersFunctioncnclDataTypeConvert
include/bang/bang_common.h:67
↓ 3 callersMethodcompare
src/nnet/Visitor/CompareMultiFormulasVisitor.cc:5
↓ 3 callersMethodcomputeFuncTune
Find the optimal computing function by comparing its running time
include/core/kernel.h:64
↓ 3 callersMethodfreeData
include/core/tensor_base.h:38
↓ 3 callersMethodgetAnchorByPattern
include/core/graph_match.h:55
↓ 3 callersMethodgetComputeOps
src/core/graph.cc:310
↓ 3 callersMethodgetDType
HACK: set correct data type
include/core/operator.h:92
↓ 3 callersMethodgetDh
include/operators/conv.h:105
↓ 3 callersMethodgetDw
include/operators/conv.h:106
↓ 3 callersMethodgetEnds
include/operators/slice.h:46
↓ 3 callersMethodgetK
include/operators/G2BMM.h:50
↓ 3 callersMethodgetNewVar
src/nnet/Pass/Pass.cc:27
↓ 3 callersMethodgetNnetExpr
include/operators/membound.h:29
↓ 3 callersMethodgetPh
include/operators/conv.h:107
↓ 3 callersMethodgetPtr
src/core/lazy_allocator.cc:170
↓ 3 callersMethodgetPw
include/operators/conv.h:108
↓ 3 callersMethodgetRoot
include/operators/broadcast.h:35
↓ 3 callersMethodgetSh
include/operators/conv.h:109
↓ 3 callersMethodgetSteps
include/operators/slice.h:52
↓ 3 callersMethodgetStmts
src/nnet/Visitor/AsTVMVisitor.cc:169
↓ 3 callersMethodgetStridesInTensor
src/nnet/iterator_table.cc:94
↓ 3 callersMethodgetSubscript
include/nnet/iterator_table.h:107
↓ 3 callersMethodgetSw
include/operators/conv.h:110
↓ 3 callersMethodgetTensor
src/core/graph.cc:130
↓ 3 callersMethodgetTransposeDim
include/operators/transpose.h:37
↓ 3 callersMethodgetWorkspaceAlloc
include/core/workspace.h:39
↓ 3 callersMethodgetXcclComm
include/kunlun/xccl_communicator.h:50
↓ 3 callersMethodhasContained
include/core/graph_match.h:50
↓ 3 callersMethodhasLoopVar
include/nnet/expr.h:245
↓ 3 callersFunctionhashVector
include/core/hash.h:11
↓ 3 callersFunctionidToRoutineType
include/nnet/expr.h:41
↓ 3 callersMethodinterpretAllOutput
src/nnet/Visitor/Interpreter.cc:170
↓ 3 callersFunctionis_sharded
(name: str)
examples/distributed/parallel_opt.py:18
↓ 3 callersFunctionmatchConv
test/nnet/test_matchConv.cc:10
↓ 3 callersMethodneq
include/nnet/expr.h:138
↓ 3 callersMethodnewTensorName
src/nnet/Pass/Pass.cc:29
↓ 3 callersFunctionpretty_print
include/nnet/dbg.h:411
↓ 3 callersFunctionrealTest
test/nnet/test_Rule2VariableMerging.cc:30
↓ 3 callersMethodresetWorkspace
include/bang/bang_runtime.h:60
↓ 3 callersMethodrun
src/cuda/cuda_runtime.cc:101
↓ 3 callersMethodsetDataBlob
src/core/tensor.cc:190
↓ 3 callersMethodsetObject
src/nnet/expr.cc:161
↓ 3 callersMethodsetPadding
include/nnet/expr.h:171
↓ 3 callersMethodsetShape
src/core/tensor.cc:62
↓ 3 callersFunctiontest_layernormFp32
test/kernels/ascend/test_ascend_layernorm.cc:11
↓ 3 callersMethodtoReadable
include/nnet/expr.h:136
↓ 3 callersMethodtoString
src/cuda/cuda_runtime.cc:114
↓ 3 callersMethodwhat
include/utils/exception.h:19
↓ 3 callersFunctionwrefs_to_refs
include/core/ref.h:36
↓ 2 callersMethodaddTarget
include/core/tensor_base.h:58
↓ 2 callersFunctionadd_value_info_for_constants
Currently onnx.shape_inference doesn't use the shape of initializers, so add that info explicitly as ValueInfoProtos. Mutates the model.
examples/distributed/kunlun/run_pytorch.py:168
↓ 2 callersFunctionadd_value_info_for_constants
Currently onnx.shape_inference doesn't use the shape of initializers, so add that info explicitly as ValueInfoProtos. Mutates the model.
examples/distributed/bang/run_pytorch_mlu.py:170
↓ 2 callersMethodallReduceSum
src/core/graph_handler.cc:511
↓ 2 callersMethodallocBlob
src/core/runtime.cc:127
↓ 2 callersMethodapplyWrapper
src/nnet/Visitor/PatternMatcher.cc:96
↓ 2 callersFunctionbitCount
src/nnet/Pass/Rule8GuidedDLT.cc:6
↓ 2 callersFunctionbuildTConv4x4_NHWF_RSFC
Warn: F is the number of input channels, which is inversed compared with normal Conv. Our data layout: NHWF -> NHWC, RSFC Pytorch data layout: NFHW ->
test/nnet/test_TConv2gemm.cc:224
↓ 2 callersFunctioncheckExprLogSame
src/nnet/test.cc:39
↓ 2 callersMethodclone
include/core/tensor.h:126
↓ 2 callersFunctioncomputeDifference2
src/utils/validation.cc:19
↓ 2 callersMethodconvTransposed2d
src/core/graph_handler.cc:69
↓ 2 callersMethodcopyBlobToCPU
include/cuda/cuda_runtime.h:76
↓ 2 callersMethoddataMalloc
include/core/tensor_base.h:32
↓ 2 callersFunctiondelocate_index
src/utils/operator_utils.cc:81
↓ 2 callersFunctionfloat_to_fp16
src/utils/data_convert.cc:5
↓ 2 callersFunctionformat_file
(file)
scripts/format.py:11
↓ 2 callersMethodgetAlphaBetaBias
include/operators/lrn.h:17
↓ 2 callersMethodgetBMNK
include/operators/matmul.h:66
↓ 2 callersMethodgetComputeType
include/operators/matmul.h:67
← previousnext →201–300 of 2,429, ranked by callers