.. currentmodule:: torch.nn.functional
.. autosummary::
:toctree: generated
:nosignatures:
conv1d
conv2d
conv3d
conv_transpose1d
conv_transpose2d
conv_transpose3d
unfold
fold
.. autosummary::
:toctree: generated
:nosignatures:
avg_pool1d
avg_pool2d
avg_pool3d
max_pool1d
max_pool2d
max_pool3d
max_unpool1d
max_unpool2d
max_unpool3d
lp_pool1d
lp_pool2d
adaptive_max_pool1d
adaptive_max_pool2d
adaptive_max_pool3d
adaptive_avg_pool1d
adaptive_avg_pool2d
adaptive_avg_pool3d
fractional_max_pool2d
fractional_max_pool3d
Non-linear activation functions
.. autosummary::
:toctree: generated
:nosignatures:
threshold
threshold_
relu
relu_
hardtanh
hardtanh_
hardswish
relu6
elu
elu_
selu
celu
leaky_relu
leaky_relu_
prelu
rrelu
rrelu_
glu
gelu
logsigmoid
hardshrink
tanhshrink
softsign
softplus
softmin
softmax
softshrink
gumbel_softmax
log_softmax
tanh
sigmoid
hardsigmoid
silu
mish
batch_norm
group_norm
instance_norm
layer_norm
local_response_norm
normalize
.. autosummary::
:toctree: generated
:nosignatures:
linear
bilinear
.. autosummary::
:toctree: generated
:nosignatures:
dropout
alpha_dropout
feature_alpha_dropout
dropout2d
dropout3d
.. autosummary::
:toctree: generated
:nosignatures:
embedding
embedding_bag
one_hot
.. autosummary::
:toctree: generated
:nosignatures:
pairwise_distance
cosine_similarity
pdist
.. autosummary::
:toctree: generated
:nosignatures:
binary_cross_entropy
binary_cross_entropy_with_logits
poisson_nll_loss
cosine_embedding_loss
cross_entropy
ctc_loss
gaussian_nll_loss
hinge_embedding_loss
kl_div
l1_loss
mse_loss
margin_ranking_loss
multilabel_margin_loss
multilabel_soft_margin_loss
multi_margin_loss
nll_loss
huber_loss
smooth_l1_loss
soft_margin_loss
triplet_margin_loss
triplet_margin_with_distance_loss
.. autosummary::
:toctree: generated
:nosignatures:
pixel_shuffle
pixel_unshuffle
pad
interpolate
upsample
upsample_nearest
upsample_bilinear
grid_sample
affine_grid
DataParallel functions (multi-GPU, distributed)
.. autosummary::
:toctree: generated
:nosignatures:
torch.nn.parallel.data_parallel