InstanceNorm1d — PyTorch 2.7 documentation (original) (raw)

class torch.nn.InstanceNorm1d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]¶

Applies Instance Normalization.

This operation applies Instance Normalization over a 2D (unbatched) or 3D (batched) input as described in the paperInstance Normalization: The Missing Ingredient for Fast Stylization.

y=x−E[x]Var[x]+ϵ∗γ+βy = \frac{x - \mathrm{E}[x]}{ \sqrt{\mathrm{Var}[x] + \epsilon}} * \gamma + \beta

The mean and standard-deviation are calculated per-dimension separately for each object in a mini-batch. γ\gamma and β\beta are learnable parameter vectors of size C (where C is the number of features or channels of the input) if affine is True. The variance is calculated via the biased estimator, equivalent totorch.var(input, unbiased=False).

By default, this layer uses instance statistics computed from input data in both training and evaluation modes.

If track_running_stats is set to True, during training this layer keeps running estimates of its computed mean and variance, which are then used for normalization during evaluation. The running estimates are kept with a default momentum of 0.1.

Note

This momentum argument is different from one used in optimizer classes and the conventional notion of momentum. Mathematically, the update rule for running statistics here isx^new=(1−momentum)×x^+momentum×xt\hat{x}_\text{new} = (1 - \text{momentum}) \times \hat{x} + \text{momentum} \times x_t, where x^\hat{x} is the estimated statistic and xtx_t is the new observed value.

Note

InstanceNorm1d and LayerNorm are very similar, but have some subtle differences. InstanceNorm1d is applied on each channel of channeled data like multidimensional time series, butLayerNorm is usually applied on entire sample and often in NLP tasks. Additionally, LayerNorm applies elementwise affine transform, while InstanceNorm1d usually don’t apply affine transform.

Parameters

• num_features (int) – number of features or channels CC of the input
• eps (float) – a value added to the denominator for numerical stability. Default: 1e-5
• momentum (Optional_[_float]) – the value used for the running_mean and running_var computation. Default: 0.1
• affine (bool) – a boolean value that when set to True, this module has learnable affine parameters, initialized the same way as done for batch normalization. Default: False.
• track_running_stats (bool) – a boolean value that when set to True, this module tracks the running mean and variance, and when set to False, this module does not track such statistics and always uses batch statistics in both training and eval modes. Default: False

Shape:

• Input: (N,C,L)(N, C, L) or (C,L)(C, L)
• Output: (N,C,L)(N, C, L) or (C,L)(C, L) (same shape as input)

Examples:

Without Learnable Parameters

m = nn.InstanceNorm1d(100)

With Learnable Parameters

m = nn.InstanceNorm1d(100, affine=True) input = torch.randn(20, 100, 40) output = m(input)