avalanche.models.packnet.PackNetModel

class avalanche.models.packnet.PackNetModel(wrappee: Module)

PackNet implements the PackNet algorithm for parameter isolation. It is designed to automatically upgrade most models to support PackNet. But because of the nature of the strategy, it is not possible to use it with every model or PyTorch module. Furthermore, PackNet not everything has been implemented yet. Here are some basic guidelines:

• Stateless modules like torch.nn.ReLU, torch.nn.Flatten,

  or torch.nn.Dropout should work fine.

• Many normalization layers currently do not work.

• Supports: nn.Linear, nn.Conv1d, nn.Conv2d,

  nn.Conv3d, nn.ConvTranspose1d, nn.ConvTranspose2d, nn.ConvTranspose3d

• If you want to use a custom module with state or parameters, ensure it

  implements PackNetModule.

Mallya, A., & Lazebnik, S. (2018). PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 7765-7773. https://doi.org/10.1109/CVPR.2018.00810

__init__(wrappee: Module) → None

Wrap a PyTorch module to make it PackNet compatible.

Parameters:

wrappee – The module to wrap

Methods

__init__(wrappee)	Wrap a PyTorch module to make it PackNet compatible.
activate_task(task_id)	Activates a task-specific subset of PackNet.
adaptation(experience)	Adapt the module (freeze units, add units...) using the current data.
add_module(name, module)	Add a child module to the current module.
apply(fn)	Apply fn recursively to every submodule (as returned by .children()) as well as self.
bfloat16()	Casts all floating point parameters and buffers to bfloat16 datatype.
buffers([recurse])	Return an iterator over module buffers.
children()	Return an iterator over immediate children modules.
compile(*args, **kwargs)	Compile this Module's forward using torch.compile().
cpu()	Move all model parameters and buffers to the CPU.
cuda([device])	Move all model parameters and buffers to the GPU.
double()	Casts all floating point parameters and buffers to double datatype.
eval()	Set the module in evaluation mode.
extra_repr()	Set the extra representation of the module.
float()	Casts all floating point parameters and buffers to float datatype.
forward(input, task_id)	Define the computation performed at every call.
forward_all_tasks(x)	compute the output given the input x and task label.
forward_single_task(x, task_label)	compute the output given the input x and task label.
freeze_pruned()	Freeze the pruned parameters, commiting them to become immutable.
get_buffer(target)	Return the buffer given by target if it exists, otherwise throw an error.
get_extra_state()	Return any extra state to include in the module's state_dict.
get_parameter(target)	Return the parameter given by target if it exists, otherwise throw an error.
get_submodule(target)	Return the submodule given by target if it exists, otherwise throw an error.
half()	Casts all floating point parameters and buffers to half datatype.
ipu([device])	Move all model parameters and buffers to the IPU.
load_state_dict(state_dict[, strict, assign])	Copy parameters and buffers from state_dict into this module and its descendants.
modules()	Return an iterator over all modules in the network.
named_buffers([prefix, recurse, ...])	Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
named_children()	Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
named_modules([memo, prefix, remove_duplicate])	Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
named_parameters([prefix, recurse, ...])	Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
parameters([recurse])	Return an iterator over module parameters.
pre_adapt(agent, experience)	Calls self.adaptation recursively accross the hierarchy of pytorch module childrens
prune(prune_proportion)	Prune a proportion of the unfrozen parameters from the module.
register_backward_hook(hook)	Register a backward hook on the module.
register_buffer(name, tensor[, persistent])	Add a buffer to the module.
register_forward_hook(hook, *[, prepend, ...])	Register a forward hook on the module.
register_forward_pre_hook(hook, *[, ...])	Register a forward pre-hook on the module.
register_full_backward_hook(hook[, prepend])	Register a backward hook on the module.
register_full_backward_pre_hook(hook[, prepend])	Register a backward pre-hook on the module.
register_load_state_dict_post_hook(hook)	Register a post hook to be run after module's load_state_dict is called.
register_module(name, module)	Alias for add_module().
register_parameter(name, param)	Add a parameter to the module.
register_state_dict_pre_hook(hook)	Register a pre-hook for the state_dict() method.
requires_grad_([requires_grad])	Change if autograd should record operations on parameters in this module.
set_extra_state(state)	Set extra state contained in the loaded state_dict.
share_memory()	See torch.Tensor.share_memory_().
state_dict(*args[, destination, prefix, ...])	Return a dictionary containing references to the whole state of the module.
to(*args, **kwargs)	Move and/or cast the parameters and buffers.
to_empty(*, device[, recurse])	Move the parameters and buffers to the specified device without copying storage.
train([mode])	Set the module in training mode.
type(dst_type)	Casts all parameters and buffers to dst_type.
wrap(wrappee)	Upgrade a PyTorch module and all of its submodules to be PackNet compatible.
xpu([device])	Move all model parameters and buffers to the XPU.
zero_grad([set_to_none])	Reset gradients of all model parameters.

Attributes

T_destination
active_task	Returns the id of the task that is currently active.
call_super_init
dump_patches
state
task_count	Counts the number of task-specific subsets in PackNet.
training
known_train_tasks_labels	Set of task labels encountered up to now.