Hybrid Module

The Hybrid module enables customization of pre-trained torchvision models. This is the star feature of v2.1.

Overview

The Hybrid module provides:

  • HybridBuilder: Load and customize pre-trained models

  • Weight Utilities: Smart weight loading with mismatch tolerance

  • Backbone Extraction: Analyze and decompose model structures

HybridBuilder

class HybridBuilder[source]

Bases: object

Builder for creating hybrid models from pre-trained backbones.

The HybridBuilder provides a fluent API for: - Loading pre-trained torchvision models - Extracting and customizing model tiers - Applying patches (attention, custom blocks) - Replacing the classification head

Example

>>> builder = HybridBuilder()
>>>
>>> # Quick build
>>> model = builder.from_torchvision(
...     "resnet50",
...     weights="IMAGENET1K_V2",
...     num_classes=100
... )
>>>
>>> # With patches
>>> model = builder.from_torchvision(
...     "efficientnet_b4",
...     weights="IMAGENET1K_V1",
...     patches={
...         "features.5": {"wrap": "SEBlock"},
...     },
...     num_classes=10,
...     dropout=0.3,
... )
SUPPORTED_BACKBONES = ['resnet18', 'resnet34', 'resnet50', 'resnet101', 'resnet152', 'wide_resnet50_2', 'wide_resnet101_2', 'resnext50_32x4d', 'resnext101_32x8d', 'resnext101_64x4d', 'efficientnet_b0', 'efficientnet_b1', 'efficientnet_b2', 'efficientnet_b3', 'efficientnet_b4', 'efficientnet_b5', 'efficientnet_b6', 'efficientnet_b7', 'efficientnet_v2_s', 'efficientnet_v2_m', 'efficientnet_v2_l', 'vgg11', 'vgg13', 'vgg16', 'vgg19', 'vgg11_bn', 'vgg13_bn', 'vgg16_bn', 'vgg19_bn', 'densenet121', 'densenet169', 'densenet201', 'densenet161', 'mobilenet_v2', 'mobilenet_v3_small', 'mobilenet_v3_large', 'convnext_tiny', 'convnext_small', 'convnext_base', 'convnext_large', 'vit_b_16', 'vit_b_32', 'vit_l_16', 'vit_l_32', 'swin_t', 'swin_s', 'swin_b']
property supported_backbones: List[str]

List of supported backbone names.

from_torchvision(backbone_name: str, weights: str | None = 'DEFAULT', patches: Dict[str, Dict] | None = None, num_classes: int = 1000, dropout: float = 0.0, freeze_backbone: bool = False, unfreeze_stages: List[int] | None = None, verbose: bool = True) → HybridModel[source]

Create a hybrid model from a torchvision backbone.

Parameters:

  • backbone_name – Name of torchvision model (e.g., ‘resnet50’)

  • weights – Weights to load (‘DEFAULT’, ‘IMAGENET1K_V1’, ‘IMAGENET1K_V2’, None)

  • patches – Dictionary of patches to apply

  • num_classes – Number of output classes

  • dropout – Dropout rate for classifier

  • freeze_backbone – Whether to freeze backbone weights

  • unfreeze_stages – If freezing, which stages to keep trainable

  • verbose – Print loading information

Returns:

HybridModel instance

Example

>>> model = builder.from_torchvision(
...     "resnet50",
...     weights="IMAGENET1K_V2",
...     patches={"layer3": {"wrap": "SEBlock"}},
...     num_classes=100,
... )
from_checkpoint(checkpoint_path: str, backbone_name: str, patches: Dict[str, Dict] | None = None, num_classes: int = 1000, dropout: float = 0.0, strict: bool = False) → HybridModel[source]

Create hybrid model from a saved checkpoint.

Parameters:

  • checkpoint_path – Path to checkpoint file

  • backbone_name – Backbone architecture name

  • patches – Patches to apply

  • num_classes – Number of output classes

  • dropout – Dropout rate

  • strict – Whether to strictly match checkpoint keys

Returns:

HybridModel loaded from checkpoint

classmethod list_backbones() → List[str][source]

List all supported backbone architectures.

Basic Usage

from torchvision_customizer import HybridBuilder

builder = HybridBuilder()

# Load ResNet50 with ImageNet weights
model = builder.from_torchvision(
    "resnet50",
    weights="IMAGENET1K_V2",
    num_classes=100,
)

Adding Patches

Patches allow you to modify specific layers:

model = builder.from_torchvision(
    "resnet50",
    weights="IMAGENET1K_V2",
    patches={
        "layer3": {"wrap": "se"},           # Wrap with SE attention
        "layer4": {"wrap": "cbam_block"},   # Wrap with CBAM
    },
    num_classes=100,
)

Patch Operations

Operation

Description

wrap

Wrap layer with attention/block

inject

Inject block after layer

replace

Replace layer entirely

Freezing for Fine-tuning

model = builder.from_torchvision(
    "resnet50",
    weights="IMAGENET1K_V2",
    num_classes=10,
    freeze_backbone=True,
    unfreeze_stages=[3],  # Only train last stage + head
)

# Later, unfreeze everything
model.unfreeze_all()

Supported Backbones

from torchvision_customizer import HybridBuilder

# List all supported backbones
print(HybridBuilder.list_backbones())

ResNet Family

  • resnet18, resnet34, resnet50, resnet101, resnet152

  • wide_resnet50_2, wide_resnet101_2

  • resnext50_32x4d, resnext101_32x8d, resnext101_64x4d

EfficientNet Family

  • efficientnet_b0 through efficientnet_b7

  • efficientnet_v2_s, efficientnet_v2_m, efficientnet_v2_l

ConvNeXt Family

  • convnext_tiny, convnext_small, convnext_base, convnext_large

MobileNet Family

  • mobilenet_v2

  • mobilenet_v3_small, mobilenet_v3_large

Other

  • VGG (11, 13, 16, 19 with/without BN)

  • DenseNet (121, 169, 201, 161)

  • Vision Transformer (vit_b_16, vit_b_32, vit_l_16, vit_l_32)

  • Swin Transformer (swin_t, swin_s, swin_b)

Weight Utilities

partial_load

partial_load(model: Module, state_dict: Dict[str, Tensor], ignore_mismatch: bool = True, strict: bool = False, verbose: bool = True, init_new_layers: str = 'kaiming') → WeightLoadingReport[source]

Load weights with tolerance for mismatches.

Handles shape mismatches, missing keys, and unexpected keys gracefully. New layers (not in state_dict) are initialized with specified method.

Parameters:

  • model – Target model to load weights into

  • state_dict – Source state dictionary

  • ignore_mismatch – If True, skip mismatched shapes instead of error

  • strict – If True, raise error on any mismatch

  • verbose – If True, print loading report

  • init_new_layers – Initialization for new layers (‘kaiming’, ‘xavier’, ‘zero’)

Returns:

WeightLoadingReport with loading statistics

Example

>>> from torchvision.models import resnet50, ResNet50_Weights
>>> base = resnet50(weights=ResNet50_Weights.IMAGENET1K_V2)
>>> custom_model = MyCustomResNet()
>>> report = partial_load(custom_model, base.state_dict())
>>> print(report.summary())

Load weights with tolerance for shape mismatches:

from torchvision_customizer import partial_load

report = partial_load(
    model,
    state_dict,
    ignore_mismatch=True,
    init_new_layers="kaiming",
)

print(report.summary())
# Loaded: 95% of parameters
# Shape Mismatch: 12 parameters
# Newly Initialized: 5 parameters

transfer_weights

transfer_weights(source: Module, target: Module, layer_mapping: Dict[str, str] | None = None, include_patterns: List[str] | None = None, exclude_patterns: List[str] | None = None) → WeightLoadingReport[source]

Transfer weights from source model to target model.

Parameters:

  • source – Source model (e.g., pretrained)

  • target – Target model (e.g., customized)

  • layer_mapping – Optional mapping from source layer names to target

  • include_patterns – Only transfer layers matching these patterns

  • exclude_patterns – Skip layers matching these patterns

Returns:

WeightLoadingReport with transfer statistics

Example

>>> pretrained = resnet50(weights='IMAGENET1K_V2')
>>> custom = CustomResNet()
>>> report = transfer_weights(pretrained, custom,
...     exclude_patterns=['fc', 'classifier'])

Transfer weights between models with filtering:

from torchvision_customizer import transfer_weights

transfer_weights(
    source=pretrained_model,
    target=custom_model,
    exclude_patterns=['fc', 'classifier'],
)

Backbone Extraction

extract_tiers

extract_tiers(model: Module, backbone_name: str | None = None) → Dict[str, Module | List[Module]][source]

Extract model into stem, stages, and head tiers.

Parameters:

  • model – The model to decompose

  • backbone_name – Optional name for lookup

Returns:

Dictionary with ‘stem’, ‘stages’, and ‘head’ keys

Example

>>> from torchvision.models import resnet50
>>> model = resnet50()
>>> tiers = extract_tiers(model)
>>> print(type(tiers['stem']))  # nn.Sequential
>>> print(len(tiers['stages']))  # 4

Decompose a model into stem, stages, and head:

from torchvision.models import resnet50
from torchvision_customizer import extract_tiers

model = resnet50()
tiers = extract_tiers(model, "resnet50")

print(f"Stem: {type(tiers['stem'])}")
print(f"Stages: {len(tiers['stages'])}")
print(f"Head: {type(tiers['head'])}")

get_backbone_info

get_backbone_info(model: Module, backbone_name: str | None = None) → BackboneInfo[source]

Get structural information about a backbone model.

Parameters:

  • model – The backbone model

  • backbone_name – Optional name hint for lookup

Returns:

BackboneInfo with structural details

Get structural information about a backbone:

from torchvision_customizer import get_backbone_info
from torchvision.models import resnet50

model = resnet50()
info = get_backbone_info(model, "resnet50")

print(info.summary())
# Backbone: resnet50
# Stem: conv1, bn1, relu, maxpool
# Stages: layer1, layer2, layer3, layer4
# Parameters: 25,557,032

HybridModel

class HybridModel(stem: Module, stages: ModuleList, head: Module, backbone_name: str, modifications: List[str] = None)[source]

Bases: Module

A hybrid model composed from a pre-trained backbone with custom modifications.

forward(x: Tensor) → Tensor[source]

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

forward_features(x: Tensor) → Tensor[source]

Extract features before the classification head.

get_stage_outputs(x: Tensor) → List[Tensor][source]

Get intermediate outputs from each stage (useful for FPN, etc.).

freeze_backbone(unfreeze_stages: List[int] = None) → HybridModel[source]

Freeze backbone, optionally leaving some stages unfrozen.

Parameters:

unfreeze_stages – List of stage indices to keep trainable

Returns:

Self for chaining

unfreeze_all() → HybridModel[source]

Unfreeze all parameters.

count_parameters(trainable_only: bool = False) → int[source]

Count model parameters.

explain() → str[source]

Generate human-readable model description.

T_destination = ~T_destination
add_module(name: str, module: Module | None) → None

Add a child module to the current module.

The module can be accessed as an attribute using the given name.

Parameters:

  • name (str) – name of the child module. The child module can be accessed from this module using the given name

  • module (Module) – child module to be added to the module.

apply(fn: Callable[[Module], None]) → Self

Apply fn recursively to every submodule (as returned by .children()) as well as self.

Typical use includes initializing the parameters of a model (see also torch.nn.init).

Parameters:

fn (Module -> None) – function to be applied to each submodule

Returns:

self

Return type:

Module

Example:

>>> @torch.no_grad()
>>> def init_weights(m):
>>>     print(m)
>>>     if type(m) == nn.Linear:
>>>         m.weight.fill_(1.0)
>>>         print(m.weight)
>>> net = nn.Sequential(nn.Linear(2, 2), nn.Linear(2, 2))
>>> net.apply(init_weights)
Linear(in_features=2, out_features=2, bias=True)
Parameter containing:
tensor([[1., 1.],
        [1., 1.]], requires_grad=True)
Linear(in_features=2, out_features=2, bias=True)
Parameter containing:
tensor([[1., 1.],
        [1., 1.]], requires_grad=True)
Sequential(
  (0): Linear(in_features=2, out_features=2, bias=True)
  (1): Linear(in_features=2, out_features=2, bias=True)
)
bfloat16() → Self

Casts all floating point parameters and buffers to bfloat16 datatype.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module

buffers(recurse: bool = True) → Iterator[Tensor]

Return an iterator over module buffers.

Parameters:

recurse (bool) – if True, then yields buffers of this module and all submodules. Otherwise, yields only buffers that are direct members of this module.

Yields:

torch.Tensor – module buffer

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> for buf in model.buffers():
>>>     print(type(buf), buf.size())
<class 'torch.Tensor'> (20L,)
<class 'torch.Tensor'> (20L, 1L, 5L, 5L)
call_super_init: bool = False
children() → Iterator[Module]

Return an iterator over immediate children modules.

Yields:

Module – a child module

compile(*args, **kwargs)

Compile this Module’s forward using torch.compile().

This Module’s __call__ method is compiled and all arguments are passed as-is to torch.compile().

See torch.compile() for details on the arguments for this function.

cpu() → Self

Move all model parameters and buffers to the CPU.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module

cuda(device: device | int | None = None) → Self

Move all model parameters and buffers to the GPU.

This also makes associated parameters and buffers different objects. So it should be called before constructing the optimizer if the module will live on GPU while being optimized.

Note

This method modifies the module in-place.

Parameters:

device (int, optional) – if specified, all parameters will be copied to that device

Returns:

self

Return type:

Module

double() → Self

Casts all floating point parameters and buffers to double datatype.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module

dump_patches: bool = False
eval() → Self

Set the module in evaluation mode.

This has an effect only on certain modules. See the documentation of particular modules for details of their behaviors in training/evaluation mode, i.e. whether they are affected, e.g. Dropout, BatchNorm, etc.

This is equivalent with self.train(False).

See Locally disabling gradient computation for a comparison between .eval() and several similar mechanisms that may be confused with it.

Returns:

self

Return type:

Module

extra_repr() → str

Return the extra representation of the module.

To print customized extra information, you should re-implement this method in your own modules. Both single-line and multi-line strings are acceptable.

float() → Self

Casts all floating point parameters and buffers to float datatype.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module

get_buffer(target: str) → Tensor

Return the buffer given by target if it exists, otherwise throw an error.

See the docstring for get_submodule for a more detailed explanation of this method’s functionality as well as how to correctly specify target.

Parameters:

target – The fully-qualified string name of the buffer to look for. (See get_submodule for how to specify a fully-qualified string.)

Returns:

The buffer referenced by target

Return type:

torch.Tensor

Raises:

AttributeError – If the target string references an invalid path or resolves to something that is not a buffer

get_extra_state() → Any

Return any extra state to include in the module’s state_dict.

Implement this and a corresponding set_extra_state() for your module if you need to store extra state. This function is called when building the module’s state_dict().

Note that extra state should be picklable to ensure working serialization of the state_dict. We only provide backwards compatibility guarantees for serializing Tensors; other objects may break backwards compatibility if their serialized pickled form changes.

Returns:

Any extra state to store in the module’s state_dict

Return type:

object

get_parameter(target: str) → Parameter

Return the parameter given by target if it exists, otherwise throw an error.

See the docstring for get_submodule for a more detailed explanation of this method’s functionality as well as how to correctly specify target.

Parameters:

target – The fully-qualified string name of the Parameter to look for. (See get_submodule for how to specify a fully-qualified string.)

Returns:

The Parameter referenced by target

Return type:

torch.nn.Parameter

Raises:

AttributeError – If the target string references an invalid path or resolves to something that is not an nn.Parameter

get_submodule(target: str) → Module

Return the submodule given by target if it exists, otherwise throw an error.

For example, let’s say you have an nn.Module A that looks like this:

A(
    (net_b): Module(
        (net_c): Module(
            (conv): Conv2d(16, 33, kernel_size=(3, 3), stride=(2, 2))
        )
        (linear): Linear(in_features=100, out_features=200, bias=True)
    )
)

(The diagram shows an nn.Module A. A which has a nested submodule net_b, which itself has two submodules net_c and linear. net_c then has a submodule conv.)

To check whether or not we have the linear submodule, we would call get_submodule("net_b.linear"). To check whether we have the conv submodule, we would call get_submodule("net_b.net_c.conv").

The runtime of get_submodule is bounded by the degree of module nesting in target. A query against named_modules achieves the same result, but it is O(N) in the number of transitive modules. So, for a simple check to see if some submodule exists, get_submodule should always be used.

Parameters:

target – The fully-qualified string name of the submodule to look for. (See above example for how to specify a fully-qualified string.)

Returns:

The submodule referenced by target

Return type:

torch.nn.Module

Raises:

AttributeError – If at any point along the path resulting from the target string the (sub)path resolves to a non-existent attribute name or an object that is not an instance of nn.Module.

half() → Self

Casts all floating point parameters and buffers to half datatype.

Note

This method modifies the module in-place.

Returns:

self

Return type:

Module

ipu(device: device | int | None = None) → Self

Move all model parameters and buffers to the IPU.

This also makes associated parameters and buffers different objects. So it should be called before constructing the optimizer if the module will live on IPU while being optimized.

Note

This method modifies the module in-place.

Parameters:

device (int, optional) – if specified, all parameters will be copied to that device

Returns:

self

Return type:

Module

load_state_dict(state_dict: Mapping[str, Any], strict: bool = True, assign: bool = False)

Copy parameters and buffers from state_dict into this module and its descendants.

If strict is True, then the keys of state_dict must exactly match the keys returned by this module’s state_dict() function.

Warning

If assign is True the optimizer must be created after the call to load_state_dict unless get_swap_module_params_on_conversion() is True.

Parameters:

  • state_dict (dict) – a dict containing parameters and persistent buffers.

  • strict (bool, optional) – whether to strictly enforce that the keys in state_dict match the keys returned by this module’s state_dict() function. Default: True

  • assign (bool, optional) – When set to False, the properties of the tensors in the current module are preserved whereas setting it to True preserves properties of the Tensors in the state dict. The only exception is the requires_grad field of Parameter for which the value from the module is preserved. Default: False

Returns:

  • missing_keys is a list of str containing any keys that are expected

    by this module but missing from the provided state_dict.

  • unexpected_keys is a list of str containing the keys that are not

    expected by this module but present in the provided state_dict.

Return type:

NamedTuple with missing_keys and unexpected_keys fields

Note

If a parameter or buffer is registered as None and its corresponding key exists in state_dict, load_state_dict() will raise a RuntimeError.

modules() → Iterator[Module]

Return an iterator over all modules in the network.

Yields:

Module – a module in the network

Note

Duplicate modules are returned only once. In the following example, l will be returned only once.

Example:

>>> l = nn.Linear(2, 2)
>>> net = nn.Sequential(l, l)
>>> for idx, m in enumerate(net.modules()):
...     print(idx, '->', m)

0 -> Sequential(
  (0): Linear(in_features=2, out_features=2, bias=True)
  (1): Linear(in_features=2, out_features=2, bias=True)
)
1 -> Linear(in_features=2, out_features=2, bias=True)
mtia(device: device | int | None = None) → Self

Move all model parameters and buffers to the MTIA.

This also makes associated parameters and buffers different objects. So it should be called before constructing the optimizer if the module will live on MTIA while being optimized.

Note

This method modifies the module in-place.

Parameters:

device (int, optional) – if specified, all parameters will be copied to that device

Returns:

self

Return type:

Module

named_buffers(prefix: str = '', recurse: bool = True, remove_duplicate: bool = True) → Iterator[tuple[str, Tensor]]

Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.

Parameters:

  • prefix (str) – prefix to prepend to all buffer names.

  • recurse (bool, optional) – if True, then yields buffers of this module and all submodules. Otherwise, yields only buffers that are direct members of this module. Defaults to True.

  • remove_duplicate (bool, optional) – whether to remove the duplicated buffers in the result. Defaults to True.

Yields:

(str, torch.Tensor) – Tuple containing the name and buffer

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> for name, buf in self.named_buffers():
>>>     if name in ['running_var']:
>>>         print(buf.size())
named_children() → Iterator[tuple[str, Module]]

Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.

Yields:

(str, Module) – Tuple containing a name and child module

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> for name, module in model.named_children():
>>>     if name in ['conv4', 'conv5']:
>>>         print(module)
named_modules(memo: set[Module] | None = None, prefix: str = '', remove_duplicate: bool = True)

Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.

Parameters:

  • memo – a memo to store the set of modules already added to the result

  • prefix – a prefix that will be added to the name of the module

  • remove_duplicate – whether to remove the duplicated module instances in the result or not

Yields:

(str, Module) – Tuple of name and module

Note

Duplicate modules are returned only once. In the following example, l will be returned only once.

Example:

>>> l = nn.Linear(2, 2)
>>> net = nn.Sequential(l, l)
>>> for idx, m in enumerate(net.named_modules()):
...     print(idx, '->', m)

0 -> ('', Sequential(
  (0): Linear(in_features=2, out_features=2, bias=True)
  (1): Linear(in_features=2, out_features=2, bias=True)
))
1 -> ('0', Linear(in_features=2, out_features=2, bias=True))
named_parameters(prefix: str = '', recurse: bool = True, remove_duplicate: bool = True) → Iterator[tuple[str, Parameter]]

Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.

Parameters:

  • prefix (str) – prefix to prepend to all parameter names.

  • recurse (bool) – if True, then yields parameters of this module and all submodules. Otherwise, yields only parameters that are direct members of this module.

  • remove_duplicate (bool, optional) – whether to remove the duplicated parameters in the result. Defaults to True.

Yields:

(str, Parameter) – Tuple containing the name and parameter

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> for name, param in self.named_parameters():
>>>     if name in ['bias']:
>>>         print(param.size())
parameters(recurse: bool = True) → Iterator[Parameter]

Return an iterator over module parameters.

This is typically passed to an optimizer.

Parameters:

recurse (bool) – if True, then yields parameters of this module and all submodules. Otherwise, yields only parameters that are direct members of this module.

Yields:

Parameter – module parameter

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> for param in model.parameters():
>>>     print(type(param), param.size())
<class 'torch.Tensor'> (20L,)
<class 'torch.Tensor'> (20L, 1L, 5L, 5L)
register_backward_hook(hook: Callable[[Module, tuple[Tensor, ...] | Tensor, tuple[Tensor, ...] | Tensor], None | tuple[Tensor, ...] | Tensor]) → RemovableHandle

Register a backward hook on the module.

This function is deprecated in favor of register_full_backward_hook() and the behavior of this function will change in future versions.

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_buffer(name: str, tensor: Tensor | None, persistent: bool = True) → None

Add a buffer to the module.

This is typically used to register a buffer that should not be considered a model parameter. For example, BatchNorm’s running_mean is not a parameter, but is part of the module’s state. Buffers, by default, are persistent and will be saved alongside parameters. This behavior can be changed by setting persistent to False. The only difference between a persistent buffer and a non-persistent buffer is that the latter will not be a part of this module’s state_dict.

Buffers can be accessed as attributes using given names.

Parameters:

  • name (str) – name of the buffer. The buffer can be accessed from this module using the given name

  • tensor (Tensor or None) – buffer to be registered. If None, then operations that run on buffers, such as cuda, are ignored. If None, the buffer is not included in the module’s state_dict.

  • persistent (bool) – whether the buffer is part of this module’s state_dict.

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> self.register_buffer('running_mean', torch.zeros(num_features))
register_forward_hook(hook: Callable[[T, tuple[Any, ...], Any], Any | None] | Callable[[T, tuple[Any, ...], dict[str, Any], Any], Any | None], *, prepend: bool = False, with_kwargs: bool = False, always_call: bool = False) → RemovableHandle

Register a forward hook on the module.

The hook will be called every time after forward() has computed an output.

If with_kwargs is False or not specified, the input contains only the positional arguments given to the module. Keyword arguments won’t be passed to the hooks and only to the forward. The hook can modify the output. It can modify the input inplace but it will not have effect on forward since this is called after forward() is called. The hook should have the following signature:

hook(module, args, output) -> None or modified output

If with_kwargs is True, the forward hook will be passed the kwargs given to the forward function and be expected to return the output possibly modified. The hook should have the following signature:

hook(module, args, kwargs, output) -> None or modified output
Parameters:

  • hook (Callable) – The user defined hook to be registered.

  • prepend (bool) – If True, the provided hook will be fired before all existing forward hooks on this torch.nn.Module. Otherwise, the provided hook will be fired after all existing forward hooks on this torch.nn.Module. Note that global forward hooks registered with register_module_forward_hook() will fire before all hooks registered by this method. Default: False

  • with_kwargs (bool) – If True, the hook will be passed the kwargs given to the forward function. Default: False

  • always_call (bool) – If True the hook will be run regardless of whether an exception is raised while calling the Module. Default: False

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_forward_pre_hook(hook: Callable[[T, tuple[Any, ...]], Any | None] | Callable[[T, tuple[Any, ...], dict[str, Any]], tuple[Any, dict[str, Any]] | None], *, prepend: bool = False, with_kwargs: bool = False) → RemovableHandle

Register a forward pre-hook on the module.

The hook will be called every time before forward() is invoked.

If with_kwargs is false or not specified, the input contains only the positional arguments given to the module. Keyword arguments won’t be passed to the hooks and only to the forward. The hook can modify the input. User can either return a tuple or a single modified value in the hook. We will wrap the value into a tuple if a single value is returned (unless that value is already a tuple). The hook should have the following signature:

hook(module, args) -> None or modified input

If with_kwargs is true, the forward pre-hook will be passed the kwargs given to the forward function. And if the hook modifies the input, both the args and kwargs should be returned. The hook should have the following signature:

hook(module, args, kwargs) -> None or a tuple of modified input and kwargs
Parameters:

  • hook (Callable) – The user defined hook to be registered.

  • prepend (bool) – If true, the provided hook will be fired before all existing forward_pre hooks on this torch.nn.Module. Otherwise, the provided hook will be fired after all existing forward_pre hooks on this torch.nn.Module. Note that global forward_pre hooks registered with register_module_forward_pre_hook() will fire before all hooks registered by this method. Default: False

  • with_kwargs (bool) – If true, the hook will be passed the kwargs given to the forward function. Default: False

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_full_backward_hook(hook: Callable[[Module, tuple[Tensor, ...] | Tensor, tuple[Tensor, ...] | Tensor], None | tuple[Tensor, ...] | Tensor], prepend: bool = False) → RemovableHandle

Register a backward hook on the module.

The hook will be called every time the gradients with respect to a module are computed, and its firing rules are as follows:

  1. Ordinarily, the hook fires when the gradients are computed with respect to the module inputs.

  2. If none of the module inputs require gradients, the hook will fire when the gradients are computed with respect to module outputs.

  3. If none of the module outputs require gradients, then the hooks will not fire.

The hook should have the following signature:

hook(module, grad_input, grad_output) -> tuple(Tensor) or None

The grad_input and grad_output are tuples that contain the gradients with respect to the inputs and outputs respectively. The hook should not modify its arguments, but it can optionally return a new gradient with respect to the input that will be used in place of grad_input in subsequent computations. grad_input will only correspond to the inputs given as positional arguments and all kwarg arguments are ignored. Entries in grad_input and grad_output will be None for all non-Tensor arguments.

For technical reasons, when this hook is applied to a Module, its forward function will receive a view of each Tensor passed to the Module. Similarly the caller will receive a view of each Tensor returned by the Module’s forward function.

Warning

Modifying inputs or outputs inplace is not allowed when using backward hooks and will raise an error.

Parameters:

  • hook (Callable) – The user-defined hook to be registered.

  • prepend (bool) – If true, the provided hook will be fired before all existing backward hooks on this torch.nn.Module. Otherwise, the provided hook will be fired after all existing backward hooks on this torch.nn.Module. Note that global backward hooks registered with register_module_full_backward_hook() will fire before all hooks registered by this method.

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_full_backward_pre_hook(hook: Callable[[Module, tuple[Tensor, ...] | Tensor], None | tuple[Tensor, ...] | Tensor], prepend: bool = False) → RemovableHandle

Register a backward pre-hook on the module.

The hook will be called every time the gradients for the module are computed. The hook should have the following signature:

hook(module, grad_output) -> tuple[Tensor] or None

The grad_output is a tuple. The hook should not modify its arguments, but it can optionally return a new gradient with respect to the output that will be used in place of grad_output in subsequent computations. Entries in grad_output will be None for all non-Tensor arguments.

For technical reasons, when this hook is applied to a Module, its forward function will receive a view of each Tensor passed to the Module. Similarly the caller will receive a view of each Tensor returned by the Module’s forward function.

Warning

Modifying inputs inplace is not allowed when using backward hooks and will raise an error.

Parameters:

  • hook (Callable) – The user-defined hook to be registered.

  • prepend (bool) – If true, the provided hook will be fired before all existing backward_pre hooks on this torch.nn.Module. Otherwise, the provided hook will be fired after all existing backward_pre hooks on this torch.nn.Module. Note that global backward_pre hooks registered with register_module_full_backward_pre_hook() will fire before all hooks registered by this method.

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_load_state_dict_post_hook(hook)

Register a post-hook to be run after module’s load_state_dict() is called.

It should have the following signature::

hook(module, incompatible_keys) -> None

The module argument is the current module that this hook is registered on, and the incompatible_keys argument is a NamedTuple consisting of attributes missing_keys and unexpected_keys. missing_keys is a list of str containing the missing keys and unexpected_keys is a list of str containing the unexpected keys.

The given incompatible_keys can be modified inplace if needed.

Note that the checks performed when calling load_state_dict() with strict=True are affected by modifications the hook makes to missing_keys or unexpected_keys, as expected. Additions to either set of keys will result in an error being thrown when strict=True, and clearing out both missing and unexpected keys will avoid an error.

Returns:

a handle that can be used to remove the added hook by calling handle.remove()

Return type:

torch.utils.hooks.RemovableHandle

register_load_state_dict_pre_hook(hook)

Register a pre-hook to be run before module’s load_state_dict() is called.

It should have the following signature::

hook(module, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs) -> None # noqa: B950

Parameters:

hook (Callable) – Callable hook that will be invoked before loading the state dict.

register_module(name: str, module: Module | None) → None

Alias for add_module().

register_parameter(name: str, param: Parameter | None) → None

Add a parameter to the module.

The parameter can be accessed as an attribute using given name.

Parameters:

  • name (str) – name of the parameter. The parameter can be accessed from this module using the given name

  • param (Parameter or None) – parameter to be added to the module. If None, then operations that run on parameters, such as cuda, are ignored. If None, the parameter is not included in the module’s state_dict.

register_state_dict_post_hook(hook)

Register a post-hook for the state_dict() method.

It should have the following signature::

hook(module, state_dict, prefix, local_metadata) -> None

The registered hooks can modify the state_dict inplace.

register_state_dict_pre_hook(hook)

Register a pre-hook for the state_dict() method.

It should have the following signature::

hook(module, prefix, keep_vars) -> None

The registered hooks can be used to perform pre-processing before the state_dict call is made.

requires_grad_(requires_grad: bool = True) → Self

Change if autograd should record operations on parameters in this module.

This method sets the parameters’ requires_grad attributes in-place.

This method is helpful for freezing part of the module for finetuning or training parts of a model individually (e.g., GAN training).

See Locally disabling gradient computation for a comparison between .requires_grad_() and several similar mechanisms that may be confused with it.

Parameters:

requires_grad (bool) – whether autograd should record operations on parameters in this module. Default: True.

Returns:

self

Return type:

Module

set_extra_state(state: Any) → None

Set extra state contained in the loaded state_dict.

This function is called from load_state_dict() to handle any extra state found within the state_dict. Implement this function and a corresponding get_extra_state() for your module if you need to store extra state within its state_dict.

Parameters:

state (dict) – Extra state from the state_dict

set_submodule(target: str, module: Module, strict: bool = False) → None

Set the submodule given by target if it exists, otherwise throw an error.

Note

If strict is set to False (default), the method will replace an existing submodule or create a new submodule if the parent module exists. If strict is set to True, the method will only attempt to replace an existing submodule and throw an error if the submodule does not exist.

For example, let’s say you have an nn.Module A that looks like this:

A(
    (net_b): Module(
        (net_c): Module(
            (conv): Conv2d(3, 3, 3)
        )
        (linear): Linear(3, 3)
    )
)

(The diagram shows an nn.Module A. A has a nested submodule net_b, which itself has two submodules net_c and linear. net_c then has a submodule conv.)

To override the Conv2d with a new submodule Linear, you could call set_submodule("net_b.net_c.conv", nn.Linear(1, 1)) where strict could be True or False

To add a new submodule Conv2d to the existing net_b module, you would call set_submodule("net_b.conv", nn.Conv2d(1, 1, 1)).

In the above if you set strict=True and call set_submodule("net_b.conv", nn.Conv2d(1, 1, 1), strict=True), an AttributeError will be raised because net_b does not have a submodule named conv.

Parameters:

  • target – The fully-qualified string name of the submodule to look for. (See above example for how to specify a fully-qualified string.)

  • module – The module to set the submodule to.

  • strict – If False, the method will replace an existing submodule or create a new submodule if the parent module exists. If True, the method will only attempt to replace an existing submodule and throw an error if the submodule doesn’t already exist.

Raises:

  • ValueError – If the target string is empty or if module is not an instance of nn.Module.

  • AttributeError – If at any point along the path resulting from the target string the (sub)path resolves to a non-existent attribute name or an object that is not an instance of nn.Module.

share_memory() → Self

See torch.Tensor.share_memory_().

state_dict(*args, destination=None, prefix='', keep_vars=False)

Return a dictionary containing references to the whole state of the module.

Both parameters and persistent buffers (e.g. running averages) are included. Keys are corresponding parameter and buffer names. Parameters and buffers set to None are not included.

Note

The returned object is a shallow copy. It contains references to the module’s parameters and buffers.

Warning

Currently state_dict() also accepts positional arguments for destination, prefix and keep_vars in order. However, this is being deprecated and keyword arguments will be enforced in future releases.

Warning

Please avoid the use of argument destination as it is not designed for end-users.

Parameters:

  • destination (dict, optional) – If provided, the state of module will be updated into the dict and the same object is returned. Otherwise, an OrderedDict will be created and returned. Default: None.

  • prefix (str, optional) – a prefix added to parameter and buffer names to compose the keys in state_dict. Default: ''.

  • keep_vars (bool, optional) – by default the Tensor s returned in the state dict are detached from autograd. If it’s set to True, detaching will not be performed. Default: False.

Returns:

a dictionary containing a whole state of the module

Return type:

dict

Example:

>>> # xdoctest: +SKIP("undefined vars")
>>> module.state_dict().keys()
['bias', 'weight']
to(*args, **kwargs)

Move and/or cast the parameters and buffers.

This can be called as

to(device=None, dtype=None, non_blocking=False)
to(dtype, non_blocking=False)
to(tensor, non_blocking=False)
to(memory_format=torch.channels_last)

Its signature is similar to torch.Tensor.to(), but only accepts floating point or complex dtypes. In addition, this method will only cast the floating point or complex parameters and buffers to dtype (if given). The integral parameters and buffers will be moved device, if that is given, but with dtypes unchanged. When non_blocking is set, it tries to convert/move asynchronously with respect to the host if possible, e.g., moving CPU Tensors with pinned memory to CUDA devices.

See below for examples.

Note

This method modifies the module in-place.

Parameters:

  • device (torch.device) – the desired device of the parameters and buffers in this module

  • dtype (torch.dtype) – the desired floating point or complex dtype of the parameters and buffers in this module

  • tensor (torch.Tensor) – Tensor whose dtype and device are the desired dtype and device for all parameters and buffers in this module

  • memory_format (torch.memory_format) – the desired memory format for 4D parameters and buffers in this module (keyword only argument)

Returns:

self

Return type:

Module

Examples:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> linear = nn.Linear(2, 2)
>>> linear.weight
Parameter containing:
tensor([[ 0.1913, -0.3420],
        [-0.5113, -0.2325]])
>>> linear.to(torch.double)
Linear(in_features=2, out_features=2, bias=True)
>>> linear.weight
Parameter containing:
tensor([[ 0.1913, -0.3420],
        [-0.5113, -0.2325]], dtype=torch.float64)
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_CUDA1)
>>> gpu1 = torch.device("cuda:1")
>>> linear.to(gpu1, dtype=torch.half, non_blocking=True)
Linear(in_features=2, out_features=2, bias=True)
>>> linear.weight
Parameter containing:
tensor([[ 0.1914, -0.3420],
        [-0.5112, -0.2324]], dtype=torch.float16, device='cuda:1')
>>> cpu = torch.device("cpu")
>>> linear.to(cpu)
Linear(in_features=2, out_features=2, bias=True)
>>> linear.weight
Parameter containing:
tensor([[ 0.1914, -0.3420],
        [-0.5112, -0.2324]], dtype=torch.float16)

>>> linear = nn.Linear(2, 2, bias=None).to(torch.cdouble)
>>> linear.weight
Parameter containing:
tensor([[ 0.3741+0.j,  0.2382+0.j],
        [ 0.5593+0.j, -0.4443+0.j]], dtype=torch.complex128)
>>> linear(torch.ones(3, 2, dtype=torch.cdouble))
tensor([[0.6122+0.j, 0.1150+0.j],
        [0.6122+0.j, 0.1150+0.j],
        [0.6122+0.j, 0.1150+0.j]], dtype=torch.complex128)
to_empty(*, device: int | str | device | None, recurse: bool = True) → Self

Move the parameters and buffers to the specified device without copying storage.

Parameters:

  • device (torch.device) – The desired device of the parameters and buffers in this module.

  • recurse (bool) – Whether parameters and buffers of submodules should be recursively moved to the specified device.

Returns:

self

Return type:

Module

train(mode: bool = True) → Self

Set the module in training mode.

This has an effect only on certain modules. See the documentation of particular modules for details of their behaviors in training/evaluation mode, i.e., whether they are affected, e.g. Dropout, BatchNorm, etc.

Parameters:

mode (bool) – whether to set training mode (True) or evaluation mode (False). Default: True.

Returns:

self

Return type:

Module

type(dst_type: dtype | str) → Self

Casts all parameters and buffers to dst_type.

Note

This method modifies the module in-place.

Parameters:

dst_type (type or string) – the desired type

Returns:

self

Return type:

Module

xpu(device: device | int | None = None) → Self

Move all model parameters and buffers to the XPU.

This also makes associated parameters and buffers different objects. So it should be called before constructing optimizer if the module will live on XPU while being optimized.

Note

This method modifies the module in-place.

Parameters:

device (int, optional) – if specified, all parameters will be copied to that device

Returns:

self

Return type:

Module

zero_grad(set_to_none: bool = True) → None

Reset gradients of all model parameters.

See similar function under torch.optim.Optimizer for more context.

Parameters:

set_to_none (bool) – instead of setting to zero, set the grads to None. See torch.optim.Optimizer.zero_grad() for details.

training: bool

Methods

  • forward(x): Standard forward pass

  • forward_features(x): Extract features before head

  • get_stage_outputs(x): Get intermediate stage outputs (for FPN)

  • freeze_backbone(unfreeze_stages): Freeze backbone layers

  • unfreeze_all(): Unfreeze all parameters

  • count_parameters(trainable_only): Count parameters

  • explain(): Human-readable model description

Example

model = HybridBuilder().from_torchvision("resnet50", ...)

# Get features for FPN
features = model.get_stage_outputs(x)

# Freeze for fine-tuning
model.freeze_backbone(unfreeze_stages=[3])

# Print summary
print(model.explain())