MCPcopy Create free account

hub / github.com/MegEngine/MegEngine / functions

Functions33,912 in github.com/MegEngine/MegEngine

↓ 14 callersMethodassert_is
assert this dtype is given type; throw exception on failure
dnn/include/megdnn/dtype.h:510
↓ 14 callersMethodcast
imperative/python/src/ops.cpp:501
↓ 14 callersMethodcurrent_scope
(self)
imperative/python/megengine/traced_module/module_tracer.py:150
↓ 14 callersMethoddesc
src/opr/impl/standalone/nms_cambricon.cpp:21
↓ 14 callersMethoddevice
src/custom/impl/tensor.cpp:465
↓ 14 callersFunctiondisable_observer
r"""Recursively disable ``module`` observer in QATModule through :meth:`~.Module.apply` Args: module: root module to do disable observer
imperative/python/megengine/quantization/quantize.py:281
↓ 14 callersMethodencode
imperative/src/include/megbrain/imperative/graph_builder.h:91
↓ 14 callersFunctionevaluate_polynomial
src/opr/test/basic_arith/erfinv.h:55
↓ 14 callersFunctionevaluate_polynomial
dnn/src/common/elemwise/erfinv.h:55
↓ 14 callersMethodexecute
(self, *args, **kwargs)
imperative/python/megengine/jit/tracing.py:279
↓ 14 callersMethodfree_device
src/core/impl/comp_node/comp_node.cpp:449
↓ 14 callersFunctionfunc
(x, w, b, dy)
imperative/python/test/unit/xla/functional/test_xla_normalize.py:39
↓ 14 callersFunctiongconv_param
dnn/test/rocm/chanwise_convolution.cpp:28
↓ 14 callersFunctionget_bundle
dnn/src/arm_common/conv_bias/int8/direct_nchw44_algo.cpp:30
↓ 14 callersFunctionget_const_value
(expr: Expr, fall_back: Any = None)
imperative/python/megengine/traced_module/_passes/utils.py:20
↓ 14 callersMethodget_default
lite/load_and_run/src/helpers/common.h:159
↓ 14 callersMethodget_dev_tensor
imperative/src/impl/interpreter/interpreter_impl.cpp:551
↓ 14 callersMethodget_mem_addr_alignment
get alignment requiement in bytes; guaranteed to be power of 2
src/core/include/megbrain/comp_node.h:285
↓ 14 callersMethodget_output_name
get the output name by the index in the network Args: index: the index of the output name Returns:
lite/pylite/megenginelite/network.py:802
↓ 14 callersFunctionget_winograd_args
dnn/test/common/conv_bias.cpp:109
↓ 14 callersFunctiongetter
src/opr/test/blas.cpp:232
↓ 14 callersFunctioninit_dataset
()
imperative/python/test/unit/data/test_dataloader.py:40
↓ 14 callersMethodinstance_id
! * \brief get current hashed instance ID */
src/core/include/megbrain/graph/operator_node.h:88
↓ 14 callersMethodis_available_attribute
dnn/src/cuda/conv_bias/algo.h:144
↓ 14 callersFunctionis_cudnn_supported
dnn/src/cuda/conv_bias/helper.cpp:126
↓ 14 callersMethodis_non_overlapping_strong
dnn/src/common/basic_types.cpp:338
↓ 14 callersMethodis_root
imperative/src/impl/interpreter/tensor_info.h:27
↓ 14 callersMethodload
lite/load_and_run/src/helpers/json_loader.cpp:300
↓ 14 callersMethodlocator_logical
get the logical locator that created this comp node
src/core/include/megbrain/comp_node.h:448
↓ 14 callersFunctionlog
r"""Element-wise :math:`\operatorname{log}(x)` function. Calculates an approximation to the natural (base :math:`e`) logarithm for each eleme
imperative/python/megengine/functional/elemwise.py:679
↓ 14 callersFunctionmatmuls
src/opr/impl/blas.cpp:622
↓ 14 callersFunctionmegcoreFree
dnn/src/common/megcore/public_api/device.cpp:71
↓ 14 callersFunctionmegcoreMalloc
dnn/src/common/megcore/public_api/device.cpp:64
↓ 14 callersFunctionnaive_transpose_kn_pad
dnn/src/x86/matrix_mul/common/common.h:413
↓ 14 callersFunctionname
dnn/src/cuda/batched_matrix_mul/algo.h:115
↓ 14 callersMethodoutput
src/serialization/include/megbrain/serialization/opr_registry.h:19
↓ 14 callersFunctionrecord_event
imperative/src/impl/physical_tensor.cpp:495
↓ 14 callersFunctionref
(x, y)
imperative/python/test/unit/functional/test_loss.py:29
↓ 14 callersMethodrelayout_opr
global relayout opr
dnn/src/cuda/handle.h:78
↓ 14 callersMethodreshape
lite/src/tensor.cpp:124
↓ 14 callersFunctionreverse_adaptor
src/core/include/megbrain/utils/metahelper.h:190
↓ 14 callersFunctionsc
dnn/src/cuda/softmax/opr_impl.cpp:45
↓ 14 callersMethodserialize
src/rdnn/impl/profiler.cpp:89
↓ 14 callersMethodslice
lite/src/tensor.cpp:191
↓ 14 callersMethodstart
dnn/test/common/timer.h:20
↓ 14 callersMethodsync
dnn/test/common/task_executor.cpp:192
↓ 14 callersMethodto_c
(self)
imperative/python/megengine/core/_wrap.py:28
↓ 14 callersFunctionvaddlvq_s16
dnn/src/arm_common/simd_macro/marm_neon.h:348
↓ 14 callersMethodvzero
dnn/src/arm_common/elemwise_helper/kimpl/fuse_add_relu.h:81
↓ 13 callersMethodConcat
src/opr/impl/tensor_manip.cpp:1215
↓ 13 callersMethodGetAPI
src/opr/impl/nvof/NvOFCuda.h:43
↓ 13 callersFunctionLITE_get_tensor_total_size_in_byte
lite/lite-c/src/tensor.cpp:185
↓ 13 callersFunctionLITE_make_network
lite/lite-c/src/network.cpp:209
↓ 13 callersFunctionMIDOUT_BEGIN
only enable midout for CPU, becuase CPU might be unused when some other platforms are used
dnn/src/common/handle.cpp:55
↓ 13 callersFunction_dump_and_load
(func, symbolic, keep_opr_name=True)
imperative/python/test/unit/utils/test_dump_naming.py:17
↓ 13 callersFunction_inplace
(f)
imperative/python/megengine/core/tensor/array_method.py:192
↓ 13 callersFunction_unwrap
(x)
imperative/python/megengine/core/tensor/megbrain_graph.py:554
↓ 13 callersMethodadd_inputs
(self, vals)
imperative/python/megengine/traced_module/expr.py:118
↓ 13 callersMethodadd_stream
src/core/impl/comp_node/mem_alloc/impl.cpp:286
↓ 13 callersMethodalign_axis
dnn/include/megdnn/tensor_format.h:116
↓ 13 callersMethodalignment_requirement
dnn/src/cambricon/handle.cpp:99
↓ 13 callersMethodalloc_shared
! * \brief allocate as shared_ptr that binds the deallocator */
src/core/include/megbrain/comp_node/alloc.h:48
↓ 13 callersMethodcall_with_opr
src/gopt/impl/framework.cpp:426
↓ 13 callersMethodcapacity
dnn/include/megdnn/thin/small_vector.h:105
↓ 13 callersMethodcomp_node
imperative/src/include/megbrain/imperative/physical_tensor.h:129
↓ 13 callersFunctiondeduce_layout
imperative/src/impl/ops/subtensor.cpp:85
↓ 13 callersFunctiondefault_network_io
lite/lite-c/src/network.cpp:62
↓ 13 callersMethoddestroy_range
dnn/include/megdnn/thin/small_vector.h:174
↓ 13 callersMethodenabled
src/core/impl/graph/cg_impl.cpp:138
↓ 13 callersMethodextra_info
! * \brief get associated extra info, or nullptr */
src/core/include/megbrain/exception.h:86
↓ 13 callersFunctionfmax
Maximum of half expressions. \param x first operand \param y second operand \return maximum of operands template<typename T,typename U> typename resul
dnn/include/megdnn/dtype/half.hpp:2208
↓ 13 callersFunctionfmin
Minimum of half expressions. \param x first operand \param y second operand \return minimum of operands template<typename T,typename U> typename resul
dnn/include/megdnn/dtype/half.hpp:2218
↓ 13 callersMethodformat
src/core/include/megbrain/graph/var_node.h:661
↓ 13 callersMethodget_all_output_name
lite/src/network.cpp:254
↓ 13 callersFunctionget_bundle
dnn/src/arm_common/conv_bias/f16/direct_nchw88_algo.cpp:20
↓ 13 callersMethodget_msecs
get milliseconds (one thousandth of a second)
src/core/include/megbrain/utils/timer.h:36
↓ 13 callersMethodget_opr_type
dnn/include/megdnn/oprs/nn.h:938
↓ 13 callersMethodinc_ref
imperative/src/include/megbrain/imperative/utils/local_ptr.h:21
↓ 13 callersMethodinput_waiting_spec
! * \brief set input waiting spec * * This should only be called from * SeqCompNodeOptimizerImpl::init_ready_event() or EagerEvalM
src/core/include/megbrain/graph/operator_node.h:571
↓ 13 callersFunctioninterleave_8
dnn/src/aarch64/matrix_mul/asm/common.h:1306
↓ 13 callersFunctioninvoke
src/core/impl/system.cpp:475
↓ 13 callersMethodis_overwrite_root
! * \brief whether this interval does not override any other interval */
src/core/impl/graph/var_node_mem_mgr/static_mem_alloc/impl.h:164
↓ 13 callersFunctionl2_block_helper
block_helper is used to calculate oh block size
dnn/src/fallback/conv_bias/gi/block_helper.h:5
↓ 13 callersFunctionmegcoreGetDeviceID
dnn/src/common/megcore/public_api/device.cpp:32
↓ 13 callersMethodmodules
r"""Returns an iterable for all the modules within this module, including itself.
imperative/python/megengine/module/module.py:349
↓ 13 callersMethodnr_threads
dnn/src/common/megcore/cpu/default_computing_context.cpp:19
↓ 13 callersMethodpermutation
r"""Randomly permute a sequence, or return a permuted range. If ``n`` is a multi-dimensional tensor, it is only shuffled along its first i
imperative/python/megengine/random/rng.py:594
↓ 13 callersMethodprocedural_name
The full procedural name indicates architecture, extended name, tile size, and layout.
dnn/scripts/cutlass_generator/gemm_operation.py:179
↓ 13 callersMethodregister_as_builtin
(cls, mod)
imperative/python/megengine/traced_module/module_tracer.py:128
↓ 13 callersFunctionreplace_vars
r"""Replaces vars in the graph. Args: dst: target vars representing the graph. varmap: the map that specifies how to replace the
imperative/python/megengine/utils/comp_graph_tools.py:280
↓ 13 callersMethodround
dnn/src/fallback/quantized_converter.h:66
↓ 13 callersMethodrun
dnn/src/cuda/cutlass/gemm_operation.h:162
↓ 13 callersFunctionrun_test
src/plugin/test/opr_footprint.cpp:45
↓ 13 callersFunctionrunner
(is_trace)
imperative/python/test/unit/xla/test_xla_training.py:43
↓ 13 callersMethodscale
src/custom/impl/tensor.cpp:332
↓ 13 callersMethodshare_memory_with
lite/src/tensor.cpp:209
↓ 13 callersMethodsin
dnn/include/megdnn/dtype/half.hpp:1338
↓ 13 callersMethodstart
(self)
imperative/python/megengine/utils/profiler.py:177
↓ 13 callersMethodstop
dnn/test/common/timer.h:26
↓ 13 callersFunctionsync
src/core/impl/comp_node_api.cpp:27
← previousnext →901–1,000 of 33,912, ranked by callers