intermediate/fx_profiling_tutorial
在 Google Colab 中运行
Colab
下载 Notebook
Notebook
在 GitHub 上查看
GitHub
注意
跳转到末尾 下载完整的示例代码。
(beta) 使用 FX 构建简单的 CPU 性能分析器#
创建于：2021 年 3 月 4 日 | 最后更新：2025 年 7 月 14 日 | 最后验证：未验证
作者：James Reed
在本教程中，我们将使用 FX 来完成以下操作：
以一种我们可以检查并收集代码结构和执行统计信息的方式捕获 PyTorch Python 代码。
构建一个小型类，作为简单的性能“分析器”，收集模型各部分在实际运行中的运行时统计信息。
在本教程中，我们将使用 torchvision 的 ResNet18 模型进行演示。
import torch
import torch.fx
import torchvision.models as models
rn18 = models.resnet18()
rn18.eval()
ResNet(
(conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)
(bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)
(layer1): Sequential(
(0): BasicBlock(
(conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
(1): BasicBlock(
(conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(layer2): Sequential(
(0): BasicBlock(
(conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(downsample): Sequential(
(0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(1): BasicBlock(
(conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(layer3): Sequential(
(0): BasicBlock(
(conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(downsample): Sequential(
(0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(1): BasicBlock(
(conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(layer4): Sequential(
(0): BasicBlock(
(conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(downsample): Sequential(
(0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(1): BasicBlock(
(conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace=True)
(conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
)
(avgpool): AdaptiveAvgPool2d(output_size=(1, 1))
(fc): Linear(in_features=512, out_features=1000, bias=True)
)
现在我们有了模型，我们想更深入地检查它的性能。也就是说，对于以下调用，模型的哪些部分花费的时间最长？
input = torch.randn(5, 3, 224, 224)
output = rn18(input)
回答这个问题的常见方法是遍历程序源代码，添加在程序各个点收集时间戳的代码，然后比较这些时间戳之间的差异，以查看时间戳之间的区域花费了多长时间。
这种技术当然适用于 PyTorch 代码，但如果我们不必复制代码并对其进行编辑，那就更好了，特别是对于我们尚未编写的代码（例如此 torchvision 模型）。相反，我们将使用 FX 来自动化这个“插桩”过程，而无需修改任何源代码。
首先，让我们处理一些导入（我们稍后将在代码中全部使用它们）。
import statistics, tabulate, time
from typing import Any, Dict, List
from torch.fx import Interpreter
注意
tabulate 是一个外部库，不是 PyTorch 的依赖项。我们将使用它来更轻松地可视化性能数据。请确保您已从您喜欢的 Python 包源安装了它。
使用符号跟踪捕获模型#
接下来，我们将使用 FX 的符号跟踪机制将模型的定义捕获到我们可以操作和检查的数据结构中。
traced_rn18 = torch.fx.symbolic_trace(rn18)
print(traced_rn18.graph)
graph():
%x : torch.Tensor [num_users=1] = placeholder[target=x]
%conv1 : [num_users=1] = call_module[target=conv1](args = (%x,), kwargs = {})
%bn1 : [num_users=1] = call_module[target=bn1](args = (%conv1,), kwargs = {})
%relu : [num_users=1] = call_module[target=relu](args = (%bn1,), kwargs = {})
%maxpool : [num_users=2] = call_module[target=maxpool](args = (%relu,), kwargs = {})
%layer1_0_conv1 : [num_users=1] = call_module[target=layer1.0.conv1](args = (%maxpool,), kwargs = {})
%layer1_0_bn1 : [num_users=1] = call_module[target=layer1.0.bn1](args = (%layer1_0_conv1,), kwargs = {})
%layer1_0_relu : [num_users=1] = call_module[target=layer1.0.relu](args = (%layer1_0_bn1,), kwargs = {})
%layer1_0_conv2 : [num_users=1] = call_module[target=layer1.0.conv2](args = (%layer1_0_relu,), kwargs = {})
%layer1_0_bn2 : [num_users=1] = call_module[target=layer1.0.bn2](args = (%layer1_0_conv2,), kwargs = {})
%add : [num_users=1] = call_function[target=operator.add](args = (%layer1_0_bn2, %maxpool), kwargs = {})
%layer1_0_relu_1 : [num_users=2] = call_module[target=layer1.0.relu](args = (%add,), kwargs = {})
%layer1_1_conv1 : [num_users=1] = call_module[target=layer1.1.conv1](args = (%layer1_0_relu_1,), kwargs = {})
%layer1_1_bn1 : [num_users=1] = call_module[target=layer1.1.bn1](args = (%layer1_1_conv1,), kwargs = {})
%layer1_1_relu : [num_users=1] = call_module[target=layer1.1.relu](args = (%layer1_1_bn1,), kwargs = {})
%layer1_1_conv2 : [num_users=1] = call_module[target=layer1.1.conv2](args = (%layer1_1_relu,), kwargs = {})
%layer1_1_bn2 : [num_users=1] = call_module[target=layer1.1.bn2](args = (%layer1_1_conv2,), kwargs = {})
%add_1 : [num_users=1] = call_function[target=operator.add](args = (%layer1_1_bn2, %layer1_0_relu_1), kwargs = {})
%layer1_1_relu_1 : [num_users=2] = call_module[target=layer1.1.relu](args = (%add_1,), kwargs = {})
%layer2_0_conv1 : [num_users=1] = call_module[target=layer2.0.conv1](args = (%layer1_1_relu_1,), kwargs = {})
%layer2_0_bn1 : [num_users=1] = call_module[target=layer2.0.bn1](args = (%layer2_0_conv1,), kwargs = {})
%layer2_0_relu : [num_users=1] = call_module[target=layer2.0.relu](args = (%layer2_0_bn1,), kwargs = {})
%layer2_0_conv2 : [num_users=1] = call_module[target=layer2.0.conv2](args = (%layer2_0_relu,), kwargs = {})
%layer2_0_bn2 : [num_users=1] = call_module[target=layer2.0.bn2](args = (%layer2_0_conv2,), kwargs = {})
%layer2_0_downsample_0 : [num_users=1] = call_module[target=layer2.0.downsample.0](args = (%layer1_1_relu_1,), kwargs = {})
%layer2_0_downsample_1 : [num_users=1] = call_module[target=layer2.0.downsample.1](args = (%layer2_0_downsample_0,), kwargs = {})
%add_2 : [num_users=1] = call_function[target=operator.add](args = (%layer2_0_bn2, %layer2_0_downsample_1), kwargs = {})
%layer2_0_relu_1 : [num_users=2] = call_module[target=layer2.0.relu](args = (%add_2,), kwargs = {})
%layer2_1_conv1 : [num_users=1] = call_module[target=layer2.1.conv1](args = (%layer2_0_relu_1,), kwargs = {})
%layer2_1_bn1 : [num_users=1] = call_module[target=layer2.1.bn1](args = (%layer2_1_conv1,), kwargs = {})
%layer2_1_relu : [num_users=1] = call_module[target=layer2.1.relu](args = (%layer2_1_bn1,), kwargs = {})
%layer2_1_conv2 : [num_users=1] = call_module[target=layer2.1.conv2](args = (%layer2_1_relu,), kwargs = {})
%layer2_1_bn2 : [num_users=1] = call_module[target=layer2.1.bn2](args = (%layer2_1_conv2,), kwargs = {})
%add_3 : [num_users=1] = call_function[target=operator.add](args = (%layer2_1_bn2, %layer2_0_relu_1), kwargs = {})
%layer2_1_relu_1 : [num_users=2] = call_module[target=layer2.1.relu](args = (%add_3,), kwargs = {})
%layer3_0_conv1 : [num_users=1] = call_module[target=layer3.0.conv1](args = (%layer2_1_relu_1,), kwargs = {})
%layer3_0_bn1 : [num_users=1] = call_module[target=layer3.0.bn1](args = (%layer3_0_conv1,), kwargs = {})
%layer3_0_relu : [num_users=1] = call_module[target=layer3.0.relu](args = (%layer3_0_bn1,), kwargs = {})
%layer3_0_conv2 : [num_users=1] = call_module[target=layer3.0.conv2](args = (%layer3_0_relu,), kwargs = {})
%layer3_0_bn2 : [num_users=1] = call_module[target=layer3.0.bn2](args = (%layer3_0_conv2,), kwargs = {})
%layer3_0_downsample_0 : [num_users=1] = call_module[target=layer3.0.downsample.0](args = (%layer2_1_relu_1,), kwargs = {})
%layer3_0_downsample_1 : [num_users=1] = call_module[target=layer3.0.downsample.1](args = (%layer3_0_downsample_0,), kwargs = {})
%add_4 : [num_users=1] = call_function[target=operator.add](args = (%layer3_0_bn2, %layer3_0_downsample_1), kwargs = {})
%layer3_0_relu_1 : [num_users=2] = call_module[target=layer3.0.relu](args = (%add_4,), kwargs = {})
%layer3_1_conv1 : [num_users=1] = call_module[target=layer3.1.conv1](args = (%layer3_0_relu_1,), kwargs = {})
%layer3_1_bn1 : [num_users=1] = call_module[target=layer3.1.bn1](args = (%layer3_1_conv1,), kwargs = {})
%layer3_1_relu : [num_users=1] = call_module[target=layer3.1.relu](args = (%layer3_1_bn1,), kwargs = {})
%layer3_1_conv2 : [num_users=1] = call_module[target=layer3.1.conv2](args = (%layer3_1_relu,), kwargs = {})
%layer3_1_bn2 : [num_users=1] = call_module[target=layer3.1.bn2](args = (%layer3_1_conv2,), kwargs = {})
%add_5 : [num_users=1] = call_function[target=operator.add](args = (%layer3_1_bn2, %layer3_0_relu_1), kwargs = {})
%layer3_1_relu_1 : [num_users=2] = call_module[target=layer3.1.relu](args = (%add_5,), kwargs = {})
%layer4_0_conv1 : [num_users=1] = call_module[target=layer4.0.conv1](args = (%layer3_1_relu_1,), kwargs = {})
%layer4_0_bn1 : [num_users=1] = call_module[target=layer4.0.bn1](args = (%layer4_0_conv1,), kwargs = {})
%layer4_0_relu : [num_users=1] = call_module[target=layer4.0.relu](args = (%layer4_0_bn1,), kwargs = {})
%layer4_0_conv2 : [num_users=1] = call_module[target=layer4.0.conv2](args = (%layer4_0_relu,), kwargs = {})
%layer4_0_bn2 : [num_users=1] = call_module[target=layer4.0.bn2](args = (%layer4_0_conv2,), kwargs = {})
%layer4_0_downsample_0 : [num_users=1] = call_module[target=layer4.0.downsample.0](args = (%layer3_1_relu_1,), kwargs = {})
%layer4_0_downsample_1 : [num_users=1] = call_module[target=layer4.0.downsample.1](args = (%layer4_0_downsample_0,), kwargs = {})
%add_6 : [num_users=1] = call_function[target=operator.add](args = (%layer4_0_bn2, %layer4_0_downsample_1), kwargs = {})
%layer4_0_relu_1 : [num_users=2] = call_module[target=layer4.0.relu](args = (%add_6,), kwargs = {})
%layer4_1_conv1 : [num_users=1] = call_module[target=layer4.1.conv1](args = (%layer4_0_relu_1,), kwargs = {})
%layer4_1_bn1 : [num_users=1] = call_module[target=layer4.1.bn1](args = (%layer4_1_conv1,), kwargs = {})
%layer4_1_relu : [num_users=1] = call_module[target=layer4.1.relu](args = (%layer4_1_bn1,), kwargs = {})
%layer4_1_conv2 : [num_users=1] = call_module[target=layer4.1.conv2](args = (%layer4_1_relu,), kwargs = {})
%layer4_1_bn2 : [num_users=1] = call_module[target=layer4.1.bn2](args = (%layer4_1_conv2,), kwargs = {})
%add_7 : [num_users=1] = call_function[target=operator.add](args = (%layer4_1_bn2, %layer4_0_relu_1), kwargs = {})
%layer4_1_relu_1 : [num_users=1] = call_module[target=layer4.1.relu](args = (%add_7,), kwargs = {})
%avgpool : [num_users=1] = call_module[target=avgpool](args = (%layer4_1_relu_1,), kwargs = {})
%flatten : [num_users=1] = call_function[target=torch.flatten](args = (%avgpool, 1), kwargs = {})
%fc : [num_users=1] = call_module[target=fc](args = (%flatten,), kwargs = {})
return fc
这为我们提供了 ResNet18 模型的 Graph 表示。Graph 由一系列相互连接的 Node 组成。每个 Node 代表 Python 代码中的一个调用点（无论是对函数、模块还是方法的调用），并且边（在每个节点上表示为 args 和 kwargs）表示在这些调用点之间传递的值。有关 Graph 表示和 FX 其他 API 的更多信息，请参阅 FX 文档 https://pytorch.ac.cn/docs/stable/fx.html。
创建性能分析解释器#
接下来，我们将创建一个继承自 torch.fx.Interpreter 的类。虽然 symbolic_trace 生成的 GraphModule 会编译在调用 GraphModule 时运行的 Python 代码，但运行 GraphModule 的另一种方法是逐个执行 Graph 中的每个 Node。这就是 Interpreter 提供的功能：它逐个节点地解释 Graph。
通过继承 Interpreter，我们可以覆盖各种功能并安装我们想要的性能分析行为。目标是创建一个对象，我们可以将模型传递给它，调用模型一次或多次，然后获取有关模型及其各个部分在这些运行中花费时间的统计信息。
让我们定义我们的 ProfilingInterpreter 类。
class ProfilingInterpreter(Interpreter):
def __init__(self, mod : torch.nn.Module):
# Rather than have the user symbolically trace their model,
# we're going to do it in the constructor. As a result, the
# user can pass in any ``Module`` without having to worry about
# symbolic tracing APIs
gm = torch.fx.symbolic_trace(mod)
super().__init__(gm)
# We are going to store away two things here:
#
# 1. A list of total runtimes for ``mod``. In other words, we are
# storing away the time ``mod(...)`` took each time this
# interpreter is called.
self.total_runtime_sec : List[float] = []
# 2. A map from ``Node`` to a list of times (in seconds) that
# node took to run. This can be seen as similar to (1) but
# for specific sub-parts of the model.
self.runtimes_sec : Dict[torch.fx.Node, List[float]] = {}
######################################################################
# Next, let's override our first method: ``run()``. ``Interpreter``'s ``run``
# method is the top-level entry point for execution of the model. We will
# want to intercept this so that we can record the total runtime of the
# model.
def run(self, *args) -> Any:
# Record the time we started running the model
t_start = time.time()
# Run the model by delegating back into Interpreter.run()
return_val = super().run(*args)
# Record the time we finished running the model
t_end = time.time()
# Store the total elapsed time this model execution took in the
# ``ProfilingInterpreter``
self.total_runtime_sec.append(t_end - t_start)
return return_val
######################################################################
# Now, let's override ``run_node``. ``Interpreter`` calls ``run_node`` each
# time it executes a single node. We will intercept this so that we
# can measure and record the time taken for each individual call in
# the model.
def run_node(self, n : torch.fx.Node) -> Any:
# Record the time we started running the op
t_start = time.time()
# Run the op by delegating back into Interpreter.run_node()
return_val = super().run_node(n)
# Record the time we finished running the op
t_end = time.time()
# If we don't have an entry for this node in our runtimes_sec
# data structure, add one with an empty list value.
self.runtimes_sec.setdefault(n, [])
# Record the total elapsed time for this single invocation
# in the runtimes_sec data structure
self.runtimes_sec[n].append(t_end - t_start)
return return_val
######################################################################
# Finally, we are going to define a method (one which doesn't override
# any ``Interpreter`` method) that provides us a nice, organized view of
# the data we have collected.
def summary(self, should_sort : bool = False) -> str:
# Build up a list of summary information for each node
node_summaries : List[List[Any]] = []
# Calculate the mean runtime for the whole network. Because the
# network may have been called multiple times during profiling,
# we need to summarize the runtimes. We choose to use the
# arithmetic mean for this.
mean_total_runtime = statistics.mean(self.total_runtime_sec)
# For each node, record summary statistics
for node, runtimes in self.runtimes_sec.items():
# Similarly, compute the mean runtime for ``node``
mean_runtime = statistics.mean(runtimes)
# For easier understanding, we also compute the percentage
# time each node took with respect to the whole network.
pct_total = mean_runtime / mean_total_runtime * 100
# Record the node's type, name of the node, mean runtime, and
# percent runtime.
node_summaries.append(
[node.op, str(node), mean_runtime, pct_total])
# One of the most important questions to answer when doing performance
# profiling is "Which op(s) took the longest?". We can make this easy
# to see by providing sorting functionality in our summary view
if should_sort:
node_summaries.sort(key=lambda s: s[2], reverse=True)
# Use the ``tabulate`` library to create a well-formatted table
# presenting our summary information
headers : List[str] = [
'Op type', 'Op', 'Average runtime (s)', 'Pct total runtime'
]
return tabulate.tabulate(node_summaries, headers=headers)
注意
我们使用 Python 的 time.time 函数来获取挂钟时间戳并进行比较。这不是衡量性能的最准确方法，只能提供一阶近似。在本教程中，我们仅出于演示目的使用这种简单技术。
调查 ResNet18 的性能#
我们现在可以使用 ProfilingInterpreter 来检查我们的 ResNet18 模型的性能特征；
interp = ProfilingInterpreter(rn18)
interp.run(input)
print(interp.summary(True))
Op type Op Average runtime (s) Pct total runtime
------------- --------------------- --------------------- -------------------
call_module maxpool 0.00485635 8.16969
call_module conv1 0.00463676 7.80029
call_module layer1_0_conv1 0.00338912 5.70142
call_module layer1_0_conv2 0.00322509 5.42547
call_module layer4_0_conv2 0.00317121 5.33483
call_module layer4_1_conv1 0.00294328 4.95139
call_module layer1_1_conv1 0.00290251 4.8828
call_module layer1_1_conv2 0.00289297 4.86676
call_module layer4_1_conv2 0.00289154 4.86435
call_module layer2_1_conv2 0.00269008 4.52544
call_module layer2_1_conv1 0.00249147 4.19133
call_module layer3_1_conv1 0.00236368 3.97635
call_module layer2_0_conv2 0.00229502 3.86084
call_module layer3_0_conv2 0.00229478 3.86044
call_module layer3_1_conv2 0.00209951 3.53195
call_module layer4_0_conv1 0.00189781 3.19263
call_module layer3_0_conv1 0.00145459 2.44702
call_module bn1 0.00137854 2.31907
call_module layer2_0_conv1 0.00126791 2.13297
call_module layer2_0_downsample_0 0.00077939 1.31115
call_module layer4_0_downsample_0 0.00050211 0.844684
call_module layer3_0_downsample_0 0.000460625 0.774895
call_function add 0.000433922 0.729974
call_function add_1 0.000392914 0.660987
call_module layer1_0_bn1 0.000322819 0.543068
call_module layer1_1_bn2 0.000308275 0.518602
call_module layer1_0_bn2 0.000286818 0.482505
call_module relu 0.000284195 0.478093
call_function add_3 0.000205278 0.345334
call_module fc 0.000194311 0.326884
call_module layer2_1_bn2 0.000166893 0.280759
call_module layer1_1_bn1 0.000156403 0.263111
call_module layer1_0_relu_1 0.000153542 0.258298
call_module layer2_0_downsample_1 0.000129938 0.218591
call_module avgpool 0.000120878 0.20335
call_module layer4_1_bn2 0.000114679 0.192922
call_module layer3_1_bn2 0.000114202 0.192119
call_module layer2_1_bn1 0.000109196 0.183697
call_module layer2_0_relu 9.98974e-05 0.168054
call_module layer1_0_relu 9.77516e-05 0.164445
call_module layer2_0_bn2 9.39369e-05 0.158027
call_module layer4_0_bn2 9.39369e-05 0.158027
call_module layer2_0_bn1 9.10759e-05 0.153214
call_module layer4_1_bn1 8.34465e-05 0.14038
call_module layer1_1_relu_1 8.32081e-05 0.139979
call_module layer3_0_bn2 8.32081e-05 0.139979
call_module layer1_1_relu 7.98702e-05 0.134363
call_function add_2 7.96318e-05 0.133962
call_module layer3_1_bn1 7.82013e-05 0.131556
call_function add_5 7.67708e-05 0.129149
call_module layer4_0_downsample_1 7.53403e-05 0.126743
output output 7.36713e-05 0.123935
call_module layer4_0_bn1 6.8903e-05 0.115913
call_module layer3_0_downsample_1 6.84261e-05 0.115111
call_module layer3_0_bn1 6.4373e-05 0.108293
call_function add_7 6.41346e-05 0.107892
call_function add_6 6.03199e-05 0.101474
call_function add_4 5.53131e-05 0.0930516
call_module layer4_1_relu 5.24521e-05 0.0882386
call_module layer4_0_relu 4.88758e-05 0.0822223
call_module layer2_0_relu_1 4.69685e-05 0.0790137
call_module layer2_1_relu_1 4.64916e-05 0.0782115
call_module layer4_0_relu_1 4.50611e-05 0.075805
call_module layer4_1_relu_1 4.43459e-05 0.0746017
call_module layer2_1_relu 4.19617e-05 0.0705909
call_module layer3_1_relu 3.79086e-05 0.0637724
call_module layer3_1_relu_1 3.60012e-05 0.0605638
call_module layer3_0_relu 3.55244e-05 0.0597616
call_module layer3_0_relu_1 3.52859e-05 0.0593605
call_function flatten 2.52724e-05 0.042515
placeholder x 1.81198e-05 0.0304824
这里有两点值得注意：
MaxPool2d花费的时间最多。这是一个已知问题：pytorch/pytorch#51393
结论#
正如我们所见，使用 FX，我们可以轻松地捕获 PyTorch 程序（即使是我们没有源代码的程序！）到一个机器可解释的格式，并将其用于分析，例如我们在此处进行的性能分析。FX 为处理 PyTorch 程序打开了一个令人兴奋的可能性世界。
最后，由于 FX 仍处于 beta 阶段，我们很乐意听取您在使用它方面的任何反馈。请随时使用 PyTorch 论坛（https://discuss.pytorch.org/）和问题跟踪器（pytorch/pytorch#issues）提供您可能有的任何反馈。
脚本总运行时间： (0 分钟 0.327 秒)
