pytorch_forecasting.models.xlstm.xLSTMTime

class pytorch_forecasting.models.xlstm.xLSTMTime(input_size: int, hidden_size: int, output_size: int, xlstm_type: Literal['slstm', 'mlstm'] = 'slstm', num_layers: int = 1, decomposition_kernel: int = 25, input_projection_size: int | None = None, dropout: float = 0.1, loss: Metric = SMAPE(), **kwargs)

xLSTMTime is a long‑term time series forecasting architecture built on the extended LSTM (xLSTM) design, incorporating either the scalar-memory stabilized LSTM (sLSTM) or the matrix-memory mLSTM variant. This model enhances classical LSTM by adding exponential gating and richer memory dynamics, and combines series decomposition and normalization layers to produce robust forecasts over extended horizons.

It is based on this paper: https://arxiv.org/pdf/2407.10240 and muslehal/xLSTMTime

Initialise the model.

Parameters:

• input_size (int) – Number of input continuous features per time step.

• hidden_size (int) – Hidden size of the xLSTM network; also used by batch norm / LSTM internals.

• output_size (int) – Number of output features per time step (forecast horizon).

• xlstm_type ({"slstm", "mlstm"}, default "slstm") – Specifies which xLSTM variant to use: - “slstm”: stabilized LSTM with scalar memory, - “mlstm”: matrix-memory variant for higher capacity and scalability.

• num_layers (int, default 1) – Number of recurrent layers in the sLSTM or mLSTM network.

• decomposition_kernel (int, default 25) – Kernel size for series decomposition into trend and seasonal components.

• input_projection_size (int, optional) – If specified, the encoded input (trend + seasonal) is projected to this size before being fed to the xLSTM; otherwise equals hidden_size.

• dropout (float, default 0.1) – Dropout rate applied within the recurrent layers.

• loss (pytorch_forecasting.metrics.Metric, default SMAPE()) – Loss (and evaluation metric) used during training.

__init__(input_size: int, hidden_size: int, output_size: int, xlstm_type: Literal['slstm', 'mlstm'] = 'slstm', num_layers: int = 1, decomposition_kernel: int = 25, input_projection_size: int | None = None, dropout: float = 0.1, loss: Metric = SMAPE(), **kwargs)

Initialise the model.

Parameters:

• input_size (int) – Number of input continuous features per time step.

• hidden_size (int) – Hidden size of the xLSTM network; also used by batch norm / LSTM internals.

• output_size (int) – Number of output features per time step (forecast horizon).

• xlstm_type ({"slstm", "mlstm"}, default "slstm") – Specifies which xLSTM variant to use: - “slstm”: stabilized LSTM with scalar memory, - “mlstm”: matrix-memory variant for higher capacity and scalability.

• num_layers (int, default 1) – Number of recurrent layers in the sLSTM or mLSTM network.

• decomposition_kernel (int, default 25) – Kernel size for series decomposition into trend and seasonal components.

• input_projection_size (int, optional) – If specified, the encoded input (trend + seasonal) is projected to this size before being fed to the xLSTM; otherwise equals hidden_size.

• dropout (float, default 0.1) – Dropout rate applied within the recurrent layers.

• loss (pytorch_forecasting.metrics.Metric, default SMAPE()) – Loss (and evaluation metric) used during training.

Methods

__call__(*args, **kwargs)	Call self as a function.
__delattr__(name)	Implement delattr(self, name).
__dir__()	Default dir() implementation.
__eq__(value, /)	Return self==value.
__format__(format_spec, /)	Default object formatter.
__ge__(value, /)	Return self>=value.
__getattr__(name)
__getattribute__(name, /)	Return getattr(self, name).
__getstate__()	Helper for pickle.
__gt__(value, /)	Return self>value.
__hash__()	Return hash(self).
__init_subclass__	This method is called when a class is subclassed.
__le__(value, /)	Return self<=value.
__lt__(value, /)	Return self<value.
__ne__(value, /)	Return self!=value.
__new__(*args, **kwargs)
__reduce__()	Helper for pickle.
__reduce_ex__(protocol, /)	Helper for pickle.
__repr__()	Return repr(self).
__setattr__(name, value)	Implement setattr(self, name, value).
__setstate__(state)
__sizeof__()	Size of object in memory, in bytes.
__str__()	Return str(self).
__subclasshook__	Abstract classes can override this to customize issubclass().
_apply(fn[, recurse])
_apply_batch_transfer_handler(batch[, ...])
_call_batch_hook(hook_name, *args)
_call_impl(*args, **kwargs)
_get_backward_hooks()	Return the backward hooks for use in the call function.
_get_backward_pre_hooks()
_get_name()
_load_from_state_dict(state_dict, prefix, ...)	Copy parameters and buffers from state_dict into only this module, but not its descendants.
_log_dict_through_fabric(dictionary[, logger])
_logger_supports(method)	Whether logger supports method.
_maybe_warn_non_full_backward_hook(inputs, ...)
_named_members(get_members_fn[, prefix, ...])	Help yield various names + members of modules.
_on_before_batch_transfer(batch[, ...])
_pkg()	Package for the model.
_register_load_state_dict_pre_hook(hook[, ...])	See register_load_state_dict_pre_hook() for details.
_register_state_dict_hook(hook)	Register a post-hook for the state_dict() method.
_replicate_for_data_parallel()
_save_to_state_dict(destination, prefix, ...)	Save module state to the destination dictionary.
_set_hparams(hp)
_slow_forward(*input, **kwargs)
_to_hparams_dict(hp)
_verify_is_manual_optimization(fn_name)
_wrapped_call_impl(*args, **kwargs)
add_module(name, module)	Add a child module to the current module.
all_gather(data[, group, sync_grads])	Gather tensors or collections of tensors from multiple processes.
apply(fn)	Apply fn recursively to every submodule (as returned by .children()) as well as self.
backward(loss, *args, **kwargs)	Called to perform backward on the loss returned in training_step().
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.
clip_gradients(optimizer[, ...])	Handles gradient clipping internally.
compile(*args, **kwargs)	Compile this Module's forward using torch.compile().
configure_callbacks()	Configure model-specific callbacks.
configure_gradient_clipping(optimizer[, ...])	Perform gradient clipping for the optimizer parameters.
configure_model()	Hook to create modules in a strategy and precision aware context.
configure_optimizers()	Configure optimizers.
configure_sharded_model()	Deprecated.
cpu()	See torch.nn.Module.cpu().
create_log(x, y, out, batch_idx[, ...])	Create the log used in the training and validation step.
cuda([device])	Moves all model parameters and buffers to the GPU.
decode_autoregressive(decode_one, ...[, ...])	Make predictions in auto-regressive manner.
deduce_default_output_parameters(dataset, kwargs)	Deduce default parameters for output for from_dataset() method.
double()	See torch.nn.Module.double().
eval()	Set the module in evaluation mode.
extra_repr()	Return extra information about parameters for representation/logging.
float()	See torch.nn.Module.float().
forward(x[, hidden_states])	Forward Pass for the model.
freeze()	Freeze all params for inference.
from_dataset(dataset, **kwargs)	Create model from dataset and set parameters related to covariates.
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()	See torch.nn.Module.half().
ipu([device])	Move all model parameters and buffers to the IPU.
load_from_checkpoint(checkpoint_path[, ...])	Primary way of loading a model from a checkpoint.
load_state_dict(state_dict[, strict, assign])	Copy parameters and buffers from state_dict into this module and its descendants.
log(*args, **kwargs)	See lightning.pytorch.core.lightning.LightningModule.log().
log_dict(dictionary[, prog_bar, logger, ...])	Log a dictionary of values at once.
log_gradient_flow(named_parameters)	log distribution of gradients to identify exploding / vanishing gradients
log_metrics(x, y, out[, prediction_kwargs])	Log metrics every training/validation step.
log_prediction(x, out, batch_idx, **kwargs)	Log metrics every training/validation step.
lr_scheduler_step(scheduler, metric)	Override this method to adjust the default way the Trainer calls each scheduler.
lr_schedulers()	Returns the learning rate scheduler(s) that are being used during training.
manual_backward(loss, *args, **kwargs)	Call this directly from your training_step() when doing optimizations manually.
modules([remove_duplicate])	Return an iterator over all modules in the network.
mtia([device])	Move all model parameters and buffers to the MTIA.
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.
on_after_backward()	Log gradient flow for debugging.
on_after_batch_transfer(batch, dataloader_idx)	Override to alter or apply batch augmentations to your batch after it is transferred to the device.
on_before_backward(loss)	Called before loss.backward().
on_before_batch_transfer(batch, dataloader_idx)	Override to alter or apply batch augmentations to your batch before it is transferred to the device.
on_before_optimizer_step(optimizer)	Called before optimizer.step().
on_before_zero_grad(optimizer)	Called after training_step() and before optimizer.zero_grad().
on_epoch_end(outputs)	Run at epoch end for training or validation.
on_fit_end()	Called at the very end of fit.
on_fit_start()	Called at the very beginning of fit.
on_load_checkpoint(checkpoint)	Called by Lightning to restore your model.
on_predict_batch_end(outputs, batch, batch_idx)	Called in the predict loop after the batch.
on_predict_batch_start(batch, batch_idx[, ...])	Called in the predict loop before anything happens for that batch.
on_predict_end()	Called at the end of predicting.
on_predict_epoch_end()	Called at the end of predicting.
on_predict_epoch_start()	Called at the beginning of predicting.
on_predict_model_eval()	Called when the predict loop starts.
on_predict_start()	Called at the beginning of predicting.
on_save_checkpoint(checkpoint)	Called by Lightning when saving a checkpoint to give you a chance to store anything else you might want to save.
on_test_batch_end(outputs, batch, batch_idx)	Called in the test loop after the batch.
on_test_batch_start(batch, batch_idx[, ...])	Called in the test loop before anything happens for that batch.
on_test_end()	Called at the end of testing.
on_test_epoch_end()	Called in the test loop at the very end of the epoch.
on_test_epoch_start()	Called in the test loop at the very beginning of the epoch.
on_test_model_eval()	Called when the test loop starts.
on_test_model_train()	Called when the test loop ends.
on_test_start()	Called at the beginning of testing.
on_train_batch_end(outputs, batch, batch_idx)	Called in the training loop after the batch.
on_train_batch_start(batch, batch_idx)	Called in the training loop before anything happens for that batch.
on_train_end()	Called at the end of training before logger experiment is closed.
on_train_epoch_end()	Called in the training loop at the very end of the epoch.
on_train_epoch_start()	Called in the training loop at the very beginning of the epoch.
on_train_start()	Called at the beginning of training after sanity check.
on_validation_batch_end(outputs, batch, ...)	Called in the validation loop after the batch.
on_validation_batch_start(batch, batch_idx)	Called in the validation loop before anything happens for that batch.
on_validation_end()	Called at the end of validation.
on_validation_epoch_end()	Called in the validation loop at the very end of the epoch.
on_validation_epoch_start()	Called in the validation loop at the very beginning of the epoch.
on_validation_model_eval()	Called when the validation loop starts.
on_validation_model_train()	Called when the validation loop ends.
on_validation_model_zero_grad()	Called by the training loop to release gradients before entering the validation loop.
on_validation_start()	Called at the beginning of validation.
optimizer_step(epoch, batch_idx, optimizer)	Override this method to adjust the default way the Trainer calls the optimizer.
optimizer_zero_grad(epoch, batch_idx, optimizer)	Override this method to change the default behaviour of optimizer.zero_grad().
optimizers([use_pl_optimizer])	Returns the optimizer(s) that are being used during training.
output_to_prediction(...[, n_samples])	Convert network output to rescaled and normalized prediction.
parameters([recurse])	Return an iterator over module parameters.
plot_prediction(x, out[, idx, ...])	Plot prediction of prediction vs actuals
predict(data[, mode, return_index, ...])	Run inference / prediction.
predict_dataloader()	An iterable or collection of iterables specifying prediction samples.
predict_dependency(data, variable, values[, ...])	Predict partial dependency.
predict_step(batch, batch_idx)	Step function called during predict().
prepare_data()	Use this to download and prepare data.
print(*args, **kwargs)	Prints only from process 0.
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_load_state_dict_pre_hook(hook)	Register a pre-hook to be run before 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_post_hook(hook)	Register a post-hook for the state_dict() method.
register_state_dict_pre_hook(hook)	Register a pre-hook for the state_dict() method.
remove_ignored_hparams(ignore_list)	Remove ignored hyperparameters from the stored state.
requires_grad_([requires_grad])	Change if autograd should record operations on parameters in this module.
save_hyperparameters(*args[, ignore, frame, ...])	Save arguments to hparams attribute.
set_extra_state(state)	Set extra state contained in the loaded state_dict.
set_submodule(target, module[, strict])	Set the submodule given by target if it exists, otherwise throw an error.
setup(stage)	Called at the beginning of fit (train + validate), validate, test, or predict.
share_memory()	See torch.Tensor.share_memory_().
size()	get number of parameters in model
state_dict(*args[, destination, prefix, ...])	Return a dictionary containing references to the whole state of the module.
step(x, y, batch_idx, **kwargs)	Run for each train/val step.
teardown(stage)	Called at the end of fit (train + validate), validate, test, or predict.
test_dataloader()	An iterable or collection of iterables specifying test samples.
test_step(batch, batch_idx)	Operates on a single batch of data from the test set.
to(*args, **kwargs)	See torch.nn.Module.to().
to_empty(*, device[, recurse])	Move the parameters and buffers to the specified device without copying storage.
to_network_output(**results)	Convert output into a named (and immutable) tuple.
to_onnx([file_path, input_sample])	Saves the model in ONNX format.
to_prediction(out[, use_metric])	Convert output to prediction using the loss metric.
to_quantiles(out[, use_metric])	Convert output to quantiles using the loss metric.
to_tensorrt([file_path, input_sample, ir, ...])	Export the model to ScriptModule or GraphModule using TensorRT compile backend.
to_torchscript([file_path, method, ...])	By default compiles the whole model to a torch.jit.ScriptModule.
toggle_optimizer(optimizer)	Makes sure only the gradients of the current optimizer's parameters are calculated in the training step to prevent dangling gradients in multiple-optimizer setup.
toggled_optimizer(optimizer)	Makes sure only the gradients of the current optimizer's parameters are calculated in the training step to prevent dangling gradients in multiple-optimizer setup.
train([mode])	Set the module in training mode.
train_dataloader()	An iterable or collection of iterables specifying training samples.
training_step(batch, batch_idx)	Train on batch.
transfer_batch_to_device(batch, device, ...)	Override this hook if your DataLoader returns tensors wrapped in a custom data structure.
transform_output(prediction, target_scale[, ...])	Extract prediction from network output and rescale it to real space / de-normalize it.
type(dst_type)	See torch.nn.Module.type().
unfreeze()	Unfreeze all parameters for training.
untoggle_optimizer(optimizer)	Resets the state of required gradients that were toggled with toggle_optimizer().
val_dataloader()	An iterable or collection of iterables specifying validation samples.
validation_step(batch, batch_idx)	Operates on a single batch of data from the validation set.
xpu([device])	Move all model parameters and buffers to the XPU.
zero_grad([set_to_none])	Reset gradients of all model parameters.

Attributes

CHECKPOINT_HYPER_PARAMS_KEY
CHECKPOINT_HYPER_PARAMS_NAME
CHECKPOINT_HYPER_PARAMS_SPECIAL_KEY
CHECKPOINT_HYPER_PARAMS_TYPE
T_destination
__annotations__
__dict__
__doc__
__jit_unused_properties__
__module__
__weakref__	list of weak references to the object
_compiled_call_impl
_jit_is_scripting
_version	This allows better BC support for load_state_dict().
automatic_optimization	If set to False you are responsible for calling .backward(), .step(), .zero_grad().
call_super_init
current_epoch	The current epoch in the Trainer, or 0 if not attached.
current_stage	Available inside lightning loops.
device
device_mesh	Strategies like ModelParallelStrategy will create a device mesh that can be accessed in the configure_model() hook to parallelize the LightningModule.
dtype
dump_patches
example_input_array	The example input array is a specification of what the module can consume in the forward() method.
fabric
global_rank	The index of the current process across all nodes and devices.
global_step	Total training batches seen across all epochs.
hparams	The collection of hyperparameters saved with save_hyperparameters().
hparams_initial	The collection of hyperparameters saved with save_hyperparameters().
lagged_target_positions	Positions of lagged target variable(s) in covariates.
local_rank	The index of the current process within a single node.
log_interval	Log interval depending if training or validating
logger	Reference to the logger object in the Trainer.
loggers	Reference to the list of loggers in the Trainer.
n_targets	Number of targets to forecast.
on_gpu	Returns True if this model is currently located on a GPU.
predicting
strict_loading	Determines how Lightning loads this model using .load_state_dict(..., strict=model.strict_loading).
target_names	List of targets that are predicted.
target_positions	Positions of target variable(s) in covariates.
trainer
training
_parameters
_buffers
_non_persistent_buffers_set
_backward_pre_hooks
_backward_hooks
_is_full_backward_hook
_forward_hooks
_forward_hooks_with_kwargs
_forward_hooks_always_called
_forward_pre_hooks
_forward_pre_hooks_with_kwargs
_state_dict_hooks
_load_state_dict_pre_hooks
_state_dict_pre_hooks
_load_state_dict_post_hooks
_modules