Statements (71)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:software
|
| https://www.w3.org/2000/01/rdf-schema#label |
PyTorch Lightning Trainer
|
| gptkbp:maintainedBy |
gptkb:PyTorch_Lightning_team
|
| gptkbp:parameter |
gptkb:performance
gptkb:strategy precision deterministic plugins callbacks devices profiler logger accelerator accumulate_grad_batches auto_lr_find auto_scale_batch_size auto_select_gpus check_val_every_n_epoch detect_anomaly enable_checkpointing enable_model_summary enable_progress_bar fast_dev_run flush_logs_every_n_steps gpus gradient_clip_algorithm gradient_clip_val ipus limit_predict_batches limit_test_batches limit_train_batches limit_val_batches log_every_n_steps max_epochs max_steps max_time min_epochs min_steps move_metrics_to_cpu multiple_trainloader_mode num_nodes num_processes num_sanity_val_steps overfit_batches progress_bar_refresh_rate reload_dataloaders_every_n_epochs replace_sampler_ddp resume_from_checkpoint stochastic_weight_avg sync_batchnorm tpu_cores track_grad_norm val_check_interval weights_summary |
| gptkbp:partOf |
gptkb:PyTorch_Lightning
|
| gptkbp:usedFor |
precision control
early stopping logging model validation multi-GPU training model testing checkpointing gradient accumulation TPU training automatic mixed precision handling distributed training managing training loops training machine learning models |
| gptkbp:writtenBy |
gptkb:Python
|
| gptkbp:bfsParent |
gptkb:Meta_AI
|
| gptkbp:bfsLayer |
5
|