Stats
Actions
Tags
From model-development-and-experiments
Organizes PyTorch code into LightningModules and DataModules, configures Trainers for multi-GPU/TPU and distributed training (DDP, FSDP, DeepSpeed), with callbacks and logging for scalable DL.
How this skill is triggered — by the user, by Claude, or both
Slash command
/model-development-and-experiments:pytorch-lightningThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
PyTorch Lightning is a deep learning framework that organizes PyTorch code to eliminate boilerplate while maintaining full flexibility. Automate training workflows, multi-device orchestration, and implement best practices for neural network training and scaling across multiple GPUs/TPUs.
PyTorch Lightning is a deep learning framework that organizes PyTorch code to eliminate boilerplate while maintaining full flexibility. Automate training workflows, multi-device orchestration, and implement best practices for neural network training and scaling across multiple GPUs/TPUs.
This skill should be used when:
Organize PyTorch models into six logical sections:
__init__() and setup()training_step(batch, batch_idx)validation_step(batch, batch_idx)test_step(batch, batch_idx)predict_step(batch, batch_idx)configure_optimizers()Quick template reference: See scripts/template_lightning_module.py for a complete boilerplate.
Detailed documentation: Read references/lightning_module.md for comprehensive method documentation, hooks, properties, and best practices.
The Trainer automates the training loop, device management, gradient operations, and callbacks. Key features:
Quick setup reference: See scripts/quick_trainer_setup.py for common Trainer configurations.
Detailed documentation: Read references/trainer.md for all parameters, methods, and configuration options.
Encapsulate all data processing steps in a reusable class:
prepare_data() - Download and process data (single-process)setup() - Create datasets and apply transforms (per-GPU)train_dataloader() - Return training DataLoaderval_dataloader() - Return validation DataLoadertest_dataloader() - Return test DataLoaderQuick template reference: See scripts/template_datamodule.py for a complete boilerplate.
Detailed documentation: Read references/data_module.md for method details and usage patterns.
Add custom functionality at specific training hooks without modifying your LightningModule. Built-in callbacks include:
Detailed documentation: Read references/callbacks.md for built-in callbacks and custom callback creation.
Integrate with multiple logging platforms:
Log metrics using self.log("metric_name", value) in any LightningModule method.
Detailed documentation: Read references/logging.md for logger setup and configuration.
Choose the right strategy based on model size:
Configure with: Trainer(strategy="ddp", accelerator="gpu", devices=4)
Detailed documentation: Read references/distributed_training.md for strategy comparison and configuration.
self.device instead of .cuda()self.save_hyperparameters() in __init__()self.log() for automatic aggregation across devicesseed_everything() and Trainer(deterministic=True)Trainer(fast_dev_run=True) to test with 1 batchDetailed documentation: Read references/best_practices.md for common patterns and pitfalls.
Define model:
class MyModel(L.LightningModule):
def __init__(self):
super().__init__()
self.save_hyperparameters()
self.model = YourNetwork()
def training_step(self, batch, batch_idx):
x, y = batch
loss = F.cross_entropy(self.model(x), y)
self.log("train_loss", loss)
return loss
def configure_optimizers(self):
return torch.optim.Adam(self.parameters())
Prepare data:
# Option 1: Direct DataLoaders
train_loader = DataLoader(train_dataset, batch_size=32)
# Option 2: LightningDataModule (recommended for reusability)
dm = MyDataModule(batch_size=32)
Train:
trainer = L.Trainer(max_epochs=10, accelerator="gpu", devices=2)
trainer.fit(model, train_loader) # or trainer.fit(model, datamodule=dm)
Executable Python templates for common PyTorch Lightning patterns:
template_lightning_module.py - Complete LightningModule boilerplatetemplate_datamodule.py - Complete LightningDataModule boilerplatequick_trainer_setup.py - Common Trainer configuration examplesDetailed documentation for each PyTorch Lightning component:
lightning_module.md - Comprehensive LightningModule guide (methods, hooks, properties)trainer.md - Trainer configuration and parametersdata_module.md - LightningDataModule patterns and methodscallbacks.md - Built-in and custom callbackslogging.md - Logger integrations and usagedistributed_training.md - DDP, FSDP, DeepSpeed comparison and setupbest_practices.md - Common patterns, tips, and pitfallsnpx claudepluginhub youyinnn/skills-collection --plugin model-development-and-experimentsOrganizes PyTorch code into LightningModules, configures Trainers for multi-GPU/TPU training, implements data pipelines, callbacks, and distributed strategies (DDP, FSDP, DeepSpeed).
Organizes PyTorch code into LightningModules, configures Trainers for multi-GPU/TPU, builds data pipelines and callbacks, and runs distributed training (DDP, FSDP, DeepSpeed). Use when structuring training loops or scaling neural-network training.
Provides PyTorch patterns and best practices for building robust, efficient, and reproducible training pipelines, model architectures, and data loading.