intermediate/fx_profiling_tutorial
在 Google Colab 中运行
Colab
下载 Notebook
Notebook
在 GitHub 上查看
GitHub
注意
转到末尾 下载完整的示例代码。
(beta) 使用 FX 构建简单的 CPU 性能剖析器#
创建日期：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：
以一种可以检查和收集代码结构及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“剖析器”，收集模型各部分在实际运行时的运行时统计信息。
在本教程中，我们将使用 torchvision 的 ResNet18 模型作为演示。
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，我们希望深入了解它的性能。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，添加在程序不同点收集时间戳的代码，并比较这些时间戳之间的差异，以了解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模型代码并对其进行编辑（尤其是我们未编写的代码，例如这个 torchvision 模型）会更好。相反，我们将使用 FX 来自动化此“仪器化”过程，而无需修改任何源代码。
首先，让我们导入一些必要的库（我们将在后续代码中使用所有这些库）。
import statistics, tabulate, time
from typing import Any, Dict, List
from torch.fx import Interpreter
注意
tabulate 是 PyTorch 的外部库，不是其依赖项。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一个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节点组成。每个节点代表 Python 代码中的一个调用点（无论是函数、模块还是方法），边（在每个节点上表示为 args 和 kwargs）代表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。
创建剖析解释器#
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：它逐节点解释图。
通过继承 Interpreter，我们可以覆盖各种功能并安装我们想要的剖析行为。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，然后获取关于模型及其各部分在这些运行期间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。这不是测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#
现在我们可以使用 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 0.00492525 8.3192
call_module conv1 0.00470901 7.95395
call_module layer4_0_conv2 0.00325871 5.50425
call_module layer1_0_conv1 0.00323415 5.46278
call_module layer4_1_conv1 0.00308633 5.21309
call_module layer4_1_conv2 0.00301981 5.10074
call_module layer1_0_conv2 0.00301218 5.08785
call_module layer1_1_conv2 0.00299501 5.05886
call_module layer1_1_conv1 0.0029788 5.03147
call_module layer3_0_conv2 0.00221872 3.74763
call_module layer3_1_conv2 0.00221801 3.74642
call_module layer2_1_conv2 0.00220633 3.72669
call_module layer2_0_conv2 0.00218749 3.69487
call_module layer3_1_conv1 0.00209832 3.54426
call_module layer2_1_conv1 0.00196481 3.31874
call_module layer4_0_conv1 0.00184798 3.12141
call_module layer2_0_conv1 0.00179219 3.02718
call_module bn1 0.00152278 2.57212
call_module layer3_0_conv1 0.00147629 2.49359
call_module layer2_0_downsample_0 0.000615597 1.0398
call_function add 0.000482321 0.814684
call_module layer4_0_downsample_0 0.000465631 0.786495
call_module layer3_0_downsample_0 0.000462055 0.780454
call_module layer1_0_bn2 0.000440121 0.743405
call_function add_1 0.000394583 0.666487
call_module relu 0.000365496 0.617356
call_module layer1_0_bn1 0.000317335 0.536008
call_module layer1_1_bn2 0.000295401 0.498959
call_function add_3 0.000225306 0.380562
call_module fc 0.000205994 0.347942
call_module layer2_0_bn1 0.000190973 0.322572
call_module layer1_1_bn1 0.00019002 0.320961
call_module layer4_1_bn1 0.000169992 0.287133
call_module avgpool 0.000126123 0.213034
call_module layer3_1_bn2 0.00011754 0.198537
call_module layer4_0_bn2 0.000111103 0.187663
call_function add_2 0.00010252 0.173166
call_module layer1_0_relu_1 9.58443e-05 0.16189
call_module layer4_1_bn2 9.58443e-05 0.16189
call_module layer2_0_bn2 9.27448e-05 0.156655
call_module layer1_0_relu 9.20296e-05 0.155446
call_module layer2_0_downsample_1 9.08375e-05 0.153433
call_module layer1_1_relu 8.65459e-05 0.146184
call_function add_5 8.55923e-05 0.144573
call_module layer1_1_relu_1 8.51154e-05 0.143768
call_module layer2_1_bn1 8.44002e-05 0.14256
call_module layer2_1_bn2 7.77245e-05 0.131284
call_module layer4_0_bn1 7.72476e-05 0.130478
call_module layer4_0_downsample_1 7.70092e-05 0.130076
call_module layer3_0_bn1 7.48634e-05 0.126451
call_module layer3_0_downsample_1 7.24792e-05 0.122424
call_module layer3_0_bn2 7.05719e-05 0.119202
call_module layer3_1_bn1 6.84261e-05 0.115578
call_function add_7 6.69956e-05 0.113162
call_function add_6 6.60419e-05 0.111551
call_module layer4_0_relu_1 5.91278e-05 0.0998723
call_module layer4_1_relu 5.74589e-05 0.0970534
call_function add_4 5.53131e-05 0.093429
call_module layer4_0_relu 5.4121e-05 0.0914154
call_module layer2_0_relu_1 4.91142e-05 0.0829585
call_module layer2_0_relu 4.86374e-05 0.0821531
call_module layer2_1_relu_1 4.86374e-05 0.0821531
call_module layer2_1_relu 4.48227e-05 0.0757097
placeholder x 4.22001e-05 0.0712799
call_module layer4_1_relu_1 4.22001e-05 0.0712799
call_module layer3_1_relu_1 4.05312e-05 0.0684609
call_module layer3_0_relu_1 3.93391e-05 0.0664473
call_module layer3_0_relu 3.83854e-05 0.0648365
call_module layer3_1_relu 3.83854e-05 0.0648365
call_function flatten 2.74181e-05 0.0463118
output output 1.07288e-05 0.018122
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，通过使用 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！）并将其转换为机器可解释的格式，用于分析，例如我们在此处进行的性能分析。FX 为处理 PyTorch 程序开启了激动人心的可能性。
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 0.311 秒)
