pytorch-lightning

Malicious tool definition detected

Tool: SKILL.md Description: --- name: pytorch-lightning description: High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate.

Malicious tool definition detected

Tool: references/callbacks.md [1/2] Description: # PyTorch Lightning Callbacks ## Overview Callbacks add functionality to training without modifying the LightningModule.

Tool: references/callbacks.md [2/2] Description: { 'counter': self.counter, 'history': self.history } def load_state_dict(self, state_dict): """Restore state.""" self.counter = state_dict['counter'] self.history = state_dict['history'] ``` ### 3.

Malicious tool definition detected

Tool: references/distributed.md [1/2] Description: # PyTorch Lightning Distributed Training ## Distributed Strategies Lightning supports multiple distributed strategies with a single parameter change.

Tool: references/distributed.md [2/2]

Malicious tool definition detected

Tool: references/hyperparameter-tuning.md [1/2] Description: # Hyperparameter Tuning with PyTorch Lightning ## Integration with Tuning Frameworks Lightning integrates seamlessly with popular hyperparameter tuning libraries.

Tool: references/hyperparameter-tuning.md [2/2] Description: around best (slow) best_lr = coarse_analysis.best_config["lr"] fine_config = { "lr": tune.uniform(best_lr * 0.5, best_lr * 2), "batch_size": tune.choice([16, 32, 64, 128]) } fine_analysis = tune.run(train_fn, config=fine_config, num_samples=20, max_epochs=20) ``` ### 2.