MCPcopy Create free account

hub / github.com/MegEngine/MegEngine / functions

Functions33,912 in github.com/MegEngine/MegEngine

↓ 21 callersMethodto_json_full
! * \brief dump to visualizer format */
src/plugin/include/megbrain/plugin/profiler.h:86
↓ 21 callersMethodto_string
imperative/python/src/transformation.h:78
↓ 21 callersFunctionwhere
(mask, x, y)
imperative/python/megengine/xla/rules/tensor.py:219
↓ 21 callersFunctionwrite
(var_ids, hlo_nodes)
imperative/python/megengine/xla/lower.py:120
↓ 20 callersMethod__init__
(self, vertical=True, horizontal=True)
imperative/python/megengine/module/vision.py:746
↓ 20 callersMethod__init__
(self, size=0, value=0, mask_value=0, *, order=None)
imperative/python/megengine/data/transform/vision/transform.py:344
↓ 20 callersMethodaddr_end
src/core/impl/graph/var_node_mem_mgr/static_mem_alloc/best_fit.h:14
↓ 20 callersMethodas_immutable_scalar
src/core/impl/graph/symbol_var.cpp:46
↓ 20 callersMethodas_nd
imperative/src/impl/basic_values.cpp:36
↓ 20 callersFunctioncb2
(_, grad)
imperative/python/test/unit/xla/test_xla_training.py:317
↓ 20 callersMethodempty
(self)
imperative/python/megengine/data/tools/_queue.py:173
↓ 20 callersFunctionfilter
(t: TileDescription)
dnn/scripts/cutlass_generator/generator.py:1041
↓ 20 callersFunctionfunc
(inp, dout)
imperative/python/test/unit/xla/functional/test_xla_tensor.py:26
↓ 20 callersMethodget_elem_size
lite/src/tensor.cpp:15
↓ 20 callersMethodget_workspace
dnn/src/x86/matrix_mul/algos.cpp:159
↓ 20 callersFunctionhash
src/core/include/megbrain/utils/hash.h:23
↓ 20 callersFunctionhash_pair_combine
! * \brief combine two hash values */
src/core/include/megbrain/utils/hash.h:13
↓ 20 callersMethodinfer_shape
src/opr/impl/tensor_manip.cpp:999
↓ 20 callersMethodload
r"""Loads a computing graph as a Network object. Args: model_path: file path of mge model. outspec: only load the sub
imperative/python/megengine/utils/network.py:64
↓ 20 callersMethodname
imperative/src/impl/transformations/dtype_promote.cpp:468
↓ 20 callersMethodname
src/jit/impl/fusion_pass.cpp:505
↓ 20 callersFunctionopr_format_to_tensor_formats
src/gopt/impl/global_layout_transform/utils.h:64
↓ 20 callersFunctionread
(var_ids)
imperative/python/megengine/xla/lower.py:111
↓ 20 callersMethodreplace_node
r"""Replace the Nodes in the graph. Args: repl_dict: the map {old_Node: new_Node} that specifies how to replace the Nodes.
imperative/python/megengine/traced_module/traced_module.py:1006
↓ 20 callersMethodreset
imperative/python/src/tensor.cpp:721
↓ 20 callersFunctionrun_binary_trans20_test
check that fop(gop(a, b), c) has been changed
src/gopt/test/gtrans.cpp:17
↓ 20 callersMethodset_hw
dnn/test/common/warp_perspective.h:57
↓ 20 callersFunctionsignbit
Check sign. \param arg number to check \retval true for negative number \retval false for positive number template<typename T> typename enable<bool,T>
dnn/include/megdnn/dtype/half.hpp:2672
↓ 20 callersFunctionstbi__get32be
lite/example/cpp_example/mge/cv/stb_image.h:1741
↓ 20 callersFunctionstbi__malloc
lite/example/cpp_example/mge/cv/stb_image.h:1011
↓ 20 callersMethodstride
src/custom/impl/tensor.cpp:474
↓ 20 callersMethodswap
src/core/impl/utils/mempool.cpp:11
↓ 20 callersFunctionupdate_max
src/core/include/megbrain/utils/arith_helper.h:21
↓ 19 callersFunctionLITE_destroy_tensor
lite/lite-c/src/tensor.cpp:71
↓ 19 callersMethod_flatten
Scans the module object and returns an iterable for the :class:`~.Tensor` and :class:`~.Module` attributes that agree with the ``predicate``.
imperative/python/megengine/module/module.py:153
↓ 19 callersMethodadd
src/core/impl/graph/helper.cpp:452
↓ 19 callersMethodadd_task
src/core/impl/utils/thread_pool.cpp:62
↓ 19 callersFunctioncallback
(impl)
imperative/python/megengine/utils/module_stats.py:61
↓ 19 callersMethodcompile
(self)
imperative/python/megengine/jit/tracing.py:276
↓ 19 callersFunctiondisable_fake_quant
r"""Recursively disable ``module`` fake quantization in QATModule through :meth:`~.Module.apply` Args: module: root module to do disable
imperative/python/megengine/quantization/quantize.py:271
↓ 19 callersMethoddtype
imperative/src/impl/value.cpp:57
↓ 19 callersFunctionexp
r"""Element-wise :math:`e^x` function. Calculates an approximation to the exponential function for each element :math:`x_i` of the input tens
imperative/python/megengine/functional/elemwise.py:592
↓ 19 callersMethodfinalize
imperative/python/src/grad_override.cpp:50
↓ 19 callersMethodfind
src/opr-mm/impl/megray_helper.cpp:56
↓ 19 callersFunctionget_winograd_mk_packed_args
dnn/test/common/conv_bias.cpp:179
↓ 19 callersMethodhas_comp_node_set
! * \brief whether at least one comp node has been set */
src/core/include/megbrain/graph/operator_node.h:120
↓ 19 callersFunctionlog_failed_load
dnn/atlas-stub/src/libatlas-wrap.cpp:162
↓ 19 callersMethodmake
(cls, *args, **kwargs)
imperative/python/megengine/traced_module/expr.py:333
↓ 19 callersFunctionmegcoreGetPlatform
dnn/src/common/megcore/public_api/device.cpp:25
↓ 19 callersMethodone_device
dnn/src/cuda/handle.h:54
↓ 19 callersMethodparam
dnn/src/cuda/convolution/backward_data/algo.h:133
↓ 19 callersMethodskip
src/serialization/impl/file.cpp:36
↓ 19 callersMethodto_string
dnn/src/common/conv_bias.cpp:181
↓ 19 callersMethodtrait
imperative/src/impl/op_def.cpp:121
↓ 19 callersMethodunflatten
r"""Given a list of values and a ``TreeDef``, builds a pytree. This is the inverse operation of ``tree_flatten``.
imperative/python/megengine/traced_module/pytree.py:246
↓ 19 callersFunctionvector_size
(vector)
scripts/gdb/commands.py:41
↓ 19 callersMethodzero_device
dnn/src/cuda/handle.h:53
↓ 18 callersFunctionGenerateConv2d
( conv_type, conv_kind, tile_descriptions, src_layout, flt_layout, dst_layout, dst
dnn/scripts/cutlass_generator/conv2d_operation.py:641
↓ 18 callersFunctionLITE_make_tensor
lite/lite-c/src/tensor.cpp:56
↓ 18 callersMethod_add_state
(self, param, state_name, initializer=None)
imperative/python/megengine/optimizer/optimizer.py:121
↓ 18 callersFunction_binary_elwise
(mode, rev=False)
imperative/python/megengine/core/tensor/array_method.py:129
↓ 18 callersFunction_convert_kwargs_to_args
( argspecs: Union[Callable, FullArgSpec], args, kwargs, is_bounded=False )
imperative/python/megengine/traced_module/utils.py:55
↓ 18 callersFunctionappend_zero_and_inc
dnn/src/x86/conv_bias/int8/common_helper.h:114
↓ 18 callersMethodas_list
r"""consume this iterator and return its content as a list
imperative/python/megengine/utils/network.py:729
↓ 18 callersFunctionbenchmark_impl
dnn/test/fallback/conv_bias.cpp:768
↓ 18 callersMethodclear
(cls)
imperative/python/megengine/utils/naming.py:18
↓ 18 callersMethodcontain
src/core/include/megbrain/graph/operator_node.h:453
↓ 18 callersMethodemit
=================== MakeShapeEmitter ====================*/
src/gopt/impl/global_layout_transform/reformat_emitter.cpp:68
↓ 18 callersMethodexec
dnn/src/common/multi_head_attn/proxy_forward_base.cpp:391
↓ 18 callersMethodexecute
src/serialization/test/extern_c_opr.cpp:60
↓ 18 callersMethodgetMgbAttributes
imperative/tablegen/helper.h:143
↓ 18 callersMethodget_device_id
lite/src/network.cpp:267
↓ 18 callersMethodget_value
dnn/test/common/small_vector.cpp:107
↓ 18 callersMethodmake_canonized_filter_meta
dnn/src/common/convolution3d.cpp:118
↓ 18 callersMethodmake_from_op_node
imperative/src/impl/op_def.cpp:13
↓ 18 callersMethodmiopen_handle
dnn/src/rocm/handle.h:46
↓ 18 callersMethodname
dnn/src/x86/matrix_mul/algos.h:153
↓ 18 callersMethodnodes
r"""Get the Nodes that constitute this graph. Args: recursive: whether to get the Nodes in the subgraph. Default:
imperative/python/megengine/traced_module/traced_module.py:649
↓ 18 callersFunctionnorm
r"""Calculates the norm of tensor elements over a given axis. This function is able to return different matrix norms, or one of an infinite
imperative/python/megengine/functional/math.py:901
↓ 18 callersMethodnr_workspace
dnn/src/common/utils.cpp:176
↓ 18 callersMethodonly_reset_raw_storage
src/core/impl/tensor.cpp:305
↓ 18 callersMethodoutput_size
imperative/src/impl/proxy_graph/mini_graph.h:534
↓ 18 callersMethodpackmode
dnn/src/fallback/matrix_mul/opr_impl.h:212
↓ 18 callersMethodremove
src/opr-mm/impl/megray_helper.cpp:73
↓ 18 callersFunctionset_log_level
dnn/test/gtest_main.cpp:16
↓ 18 callersMethodto_json
src/plugin/impl/profiler.cpp:143
↓ 18 callersMethodunlock
src/core/impl/utils/thread.cpp:31
↓ 18 callersMethodunwrap_inputs
imperative/src/impl/transformations/format.cpp:77
↓ 18 callersFunctionvcvtaq_s32_f32
dnn/src/arm_common/simd_macro/marm_neon.h:730
↓ 17 callersFunctionLITE_get_io_tensor
lite/lite-c/src/network.cpp:272
↓ 17 callersFunctionMGB_CATCH
src/core/impl/graph/cg_impl.cpp:103
↓ 17 callersFunction_shape_equal
(lhs_shape, rhs_shape)
imperative/python/megengine/xla/rules/utils.py:45
↓ 17 callersFunctionapply_on_physical_tensor
imperative/src/impl/ops/rng.cpp:710
↓ 17 callersMethodattach
r"""Instruct GradManager to track operations on tensors, so that gradients with respect to those tensors could be evaluated later. :m
imperative/python/megengine/autodiff/grad_manager.py:131
↓ 17 callersFunctionbenchmark_with_contrast
dnn/test/common/conv_bias.cpp:1127
↓ 17 callersMethodcompile
r"""Delete unused expr.
imperative/python/megengine/traced_module/traced_module.py:1086
↓ 17 callersFunctionconcat
(inps, axis)
imperative/python/megengine/xla/rules/tensor.py:116
↓ 17 callersFunctionconcrete_handle
dnn/src/rocm/utils.h:16
↓ 17 callersMethoddata
imperative/src/impl/transformations/trace.cpp:695
↓ 17 callersMethoddevice_wait_event
src/core/include/megbrain/comp_node.h:794
← previousnext →701–800 of 33,912, ranked by callers