noether.modeling.functional.init

Attributes

ALL_CONVS

ALL_LAYERS

Functions

init_bias_to_zero(layer_module)

Initialize the bias tensor of a nn.Module instance to zero.

init_trunc_normal_zero_bias(layer_module[, std])

Initialize the weight tensor of a nn.Module instance using the truncated normal initialization with a zero bias

apply_init_method(module, proj_weight, init_method)

Apply an initialization function to all applicable sub-modules of a given module.

Module Contents

noether.modeling.functional.init.ALL_CONVS
noether.modeling.functional.init.ALL_LAYERS
noether.modeling.functional.init.init_bias_to_zero(layer_module)

Initialize the bias tensor of a nn.Module instance to zero.

Parameters:

layer_module (torch.nn.Module) – An nn.Module instance, either a Linear or Conv layer.

Return type:

None

noether.modeling.functional.init.init_trunc_normal_zero_bias(layer_module, std=0.02)

Initialize the weight tensor of a nn.Module instance using the truncated normal initialization with a zero bias vector.

Parameters:

  • layer_module (torch.nn.Module) – An nn.Module instance, either a Linear or Conv layer.

  • std (float) – Standard Deviation value of the normal distribution to sample weights from. Defaults to 0.02.

Return type:

None

noether.modeling.functional.init.apply_init_method(module, proj_weight, init_method)

Apply an initialization function to all applicable sub-modules of a given module.

Parameters:

  • module (torch.nn.Module) – The nn.Module instance to initialize.

  • init_fn – The initialization function to apply to each sub-module.

  • proj_weight (torch.Tensor)

  • init_method (str)

Return type:

None