Flops counting tool for neural networks in pytorch framework

This tool is designed to compute the theoretical amount of multiply-add operations in neural networks. It can also compute the number of parameters and print per-layer computational cost of a given network.

ptflops has two backends, pytorch and aten. pytorch backend is a legacy one, it considers nn.Modules only. However, it's still useful, since it provides a better par-layer analytics for CNNs. In all other cases it's recommended to use aten backend, which considers aten operations, and therefore it covers more model architectures (including transformers). The default backend is aten. Please, don't use pytorch backend for transformer architectures.

aten backend

Operations considered:

• aten.mm, aten.matmul, aten.addmm, aten.bmm
• aten.convolution

Usage tips

• Use verbose=True to see the operations which were not considered during complexity computation.
• This backend prints per-module statistics only for modules directly nested into the root nn.Module. Deeper modules at the second level of nesting are not shown in the per-layer statistics.
• ignore_modules option forces ptflops to ignore the listed modules. This can be useful for research purposes. For instance, one can drop all convolutions from the counting process specifying ignore_modules=[torch.ops.aten.convolution, torch.ops.aten._convolution].

pytorch backend

Supported layers:

• Conv1d/2d/3d (including grouping)
• ConvTranspose1d/2d/3d (including grouping)
• BatchNorm1d/2d/3d, GroupNorm, InstanceNorm1d/2d/3d, LayerNorm
• Activations (ReLU, PReLU, ELU, ReLU6, LeakyReLU, GELU)
• Linear
• Upsample
• Poolings (AvgPool1d/2d/3d, MaxPool1d/2d/3d and adaptive ones)

Experimental support:

• RNN, LSTM, GRU (NLH layout is assumed)
• RNNCell, LSTMCell, GRUCell
• torch.nn.MultiheadAttention
• torchvision.ops.DeformConv2d
• visual transformers from timm

Usage tips

• This backend doesn't take into account some of the torch.nn.functional.* and tensor.* operations. Therefore unsupported operations are not contributing to the final complexity estimation. See ptflops/pytorch_ops.py:FUNCTIONAL_MAPPING,TENSOR_OPS_MAPPING to check supported ops. Sometimes functional-level hooks conflict with hooks for nn.Module (for instance, custom ones). In that case, counting with these ops can be disabled by passing backend_specific_config={"count_functional" : False}.
• ptflops launches a given model on a random tensor and estimates amount of computations during inference. Complicated models can have several inputs, some of them could be optional. To construct non-trivial input one can use the input_constructor argument of the get_model_complexity_info. input_constructor is a function that takes the input spatial resolution as a tuple and returns a dict with named input arguments of the model. Next, this dict would be passed to the model as a keyword arguments.
• verbose parameter allows to get information about modules that don't contribute to the final numbers.
• ignore_modules option forces ptflops to ignore the listed modules. This can be useful for research purposes. For instance, one can drop all convolutions from the counting process specifying ignore_modules=[torch.nn.Conv2d].

Requirements

Pytorch >= 2.0. Use pip install ptflops==0.7.2.2 to work with torch 1.x.

Install the latest version

From PyPI:

pip install ptflops

From this repository:

pip install --upgrade git+https://github.com/sovrasov/flops-counter.pytorch.git

Example

import torchvision.models as models
import torch
from ptflops import get_model_complexity_info

with torch.cuda.device(0):
  net = models.densenet161()
  macs, params = get_model_complexity_info(net, (3, 224, 224), as_strings=True, backend='pytorch'
                                           print_per_layer_stat=True, verbose=True)
  print('{:<30}  {:<8}'.format('Computational complexity: ', macs))
  print('{:<30}  {:<8}'.format('Number of parameters: ', params))

  macs, params = get_model_complexity_info(net, (3, 224, 224), as_strings=True, backend='aten'
                                           print_per_layer_stat=True, verbose=True)
  print('{:<30}  {:<8}'.format('Computational complexity: ', macs))
  print('{:<30}  {:<8}'.format('Number of parameters: ', params))

Citation

If ptflops was useful for your paper or tech report, please cite me:

@online{ptflops,
  author = {Vladislav Sovrasov},
  title = {ptflops: a flops counting tool for neural networks in pytorch framework},
  year = 2018-2024,
  url = {https://github.com/sovrasov/flops-counter.pytorch},
}

Credits

Thanks to @warmspringwinds and Horace He for the initial version of the script.

Benchmark

torchvision

Model	Input Resolution	Params(M)	MACs(G) (pytorch)	MACs(G) (aten)
alexnet	224x224	61.10	0.72	0.71
convnext_base	224x224	88.59	15.43	15.38
densenet121	224x224	7.98	2.90
efficientnet_b0	224x224	5.29	0.41
efficientnet_v2_m	224x224	54.14	5.43
googlenet	224x224	13.00	1.51
inception_v3	224x224	27.16	5.75	5.71
maxvit_t	224x224	30.92	5.48
mnasnet1_0	224x224	4.38	0.33
mobilenet_v2	224x224	3.50	0.32
mobilenet_v3_large	224x224	5.48	0.23
regnet_y_1_6gf	224x224	11.20	1.65
resnet18	224x224	11.69	1.83	1.81
resnet50	224x224	25.56	4.13	4.09
resnext50_32x4d	224x224	25.03	4.29
shufflenet_v2_x1_0	224x224	2.28	0.15
squeezenet1_0	224x224	1.25	0.84	0.82
vgg16	224x224	138.36	15.52	15.48
vit_b_16	224x224	86.57	17.61 (wrong)	16.86
wide_resnet50_2	224x224	68.88	11.45

timm

Model | Input Resolution | Params(M) | MACs(G)