PyTorch/Lightning Toolkit¶
Datasets¶
|
A dataset wrapper that caches the samples to improve performance. |
|
A dataset wrapper that caches samples to disk to reduce memory usage. |
|
A dataset class for loading data from an HDF5 file. |
|
A dataset class for reading data from HDF5 files. |
|
Splits a dataset into subsets and returns the subset corresponding to the current process rank. |
Functional¶
|
Solve the linear sum assignment problem for a batch of cost matrices. |
Learning Rate Schedulers¶
|
Build a sequential learning rate scheduler from scheduler factory functions. |
Inference Pipeline Helpers¶
|
Base class for inference pipelines. |
|
Reorder attrs fields such that fields without default values come first, then fields with default values. |
Lightning Utilities¶
|
LightningModule subclass for models that use custom optimizers and schedulers. |
|
Configuration for a learning rate scheduler in Lightning. |
|
Configuration for the optimizer and scheduler in a LightningModule. |
|
Build parameter groups for the optimizer based on the provided patterns. |
|
Load a PyTorch Lightning model from a pre-trained checkpoint. |
Environment class for ABCI. |
Lightning Datamodules¶
|
A PyTorch Lightning DataModule designed to handle training and validation datasets with support for caching and dataset scattering. |
Lightning Callbacks¶
|
Lightning Callback for save metric in fit ends. |
|
Initialize a model from a pretrained checkpoint before training or validation. |
|
Weight averaging callback that ignores buffers during averaging and swapping. |
|
Exponential moving average (EMA) callback that ignores buffers. |
Warn once when trainable parameters do not receive gradients. |
H5py Utilities¶
Abstract base class for writing data to an HDF5 file. |