class ignite.metrics.InceptionScore(num_features=None, feature_extractor=None, output_transform=<function InceptionScore.<lambda>>, device=device(type='cpu'))[source]#

Calculates Inception Score.

IS(G)=exp(1Ni=1NDKL(p(yx(i)p^(y))))\text{IS(G)} = \exp(\frac{1}{N}\sum_{i=1}^{N} D_{KL} (p(y|x^{(i)} \parallel \hat{p}(y))))

where p(yx)p(y|x) is the conditional probability of image being the given object and p(y)p(y) is the marginal probability that the given image is real, G refers to the generated image and DKLD_{KL} refers to KL Divergence of the above mentioned probabilities.

More details can be found in Barratt et al. 2018.

  • num_features (Optional[int]) – number of features predicted by the model or number of classes of the model. Default value is 1000.

  • feature_extractor (Optional[Module]) – a torch Module for predicting the probabilities from the input data. It returns a tensor of shape (batch_size, num_features). If neither num_features nor feature_extractor are defined, by default we use an ImageNet pretrained Inception Model. If only num_features is defined but feature_extractor is not defined, feature_extractor is assigned Identity Function. Please note that the class object will be implicitly converted to device mentioned in the device argument.

  • output_transform (Callable) – a callable that is used to transform the Engine’s process_function’s output into the form expected by the metric. This can be useful if, for example, you have a multi-output model and you want to compute the metric with respect to one of the outputs. By default, metrics require the output as y_pred.

  • device (Union[str, device]) – specifies which device updates are accumulated on. Setting the metric’s device to be the same as your update arguments ensures the update method is non-blocking. By default, CPU.


The default Inception model requires the torchvision module to be installed.


For more information on how metric works with Engine, visit Attach Engine API.

from collections import OrderedDict

import torch
from torch import nn, optim

from ignite.engine import *
from ignite.handlers import *
from ignite.metrics import *
from ignite.utils import *
from ignite.contrib.metrics.regression import *
from ignite.contrib.metrics import *

# create default evaluator for doctests

def eval_step(engine, batch):
    return batch

default_evaluator = Engine(eval_step)

# create default optimizer for doctests

param_tensor = torch.zeros([1], requires_grad=True)
default_optimizer = torch.optim.SGD([param_tensor], lr=0.1)

# create default trainer for doctests
# as handlers could be attached to the trainer,
# each test must define his own trainer using `.. testsetup:`

def get_default_trainer():

    def train_step(engine, batch):
        return batch

    return Engine(train_step)

# create default model for doctests

default_model = nn.Sequential(OrderedDict([
    ('base', nn.Linear(4, 2)),
    ('fc', nn.Linear(2, 1))

metric = InceptionScore()
metric.attach(default_evaluator, "is")
y = torch.rand(10, 3, 299, 299)
state =[y])
metric = InceptionScore(num_features=1, feature_extractor=default_model)
metric.attach(default_evaluator, "is")
y = torch.zeros(10, 4)
state =[y])

New in version 0.4.6.



Computes the metric based on it's accumulated state.


Resets the metric to it's initial state.


Updates the metric's state using the passed batch output.


Computes the metric based on it’s accumulated state.

By default, this is called at the end of each epoch.


the actual quantity of interest. However, if a Mapping is returned, it will be (shallow) flattened into engine.state.metrics when completed() is called.

Return type



NotComputableError – raised when the metric cannot be computed.


Resets the metric to it’s initial state.

By default, this is called at the start of each epoch.

Return type



Updates the metric’s state using the passed batch output.

By default, this is called once for each batch.


output (Tensor) – the is the output from the engine’s process function.

Return type