torchtrain

Subpackages

• callbacks
  • Submodules
    • abstract
      • AbstractCallback
      • AbstractExperimentCallback
    • basic
      • ListRegisteredCallbacks
      • EarlyStopping
      • ThresholdEarlyStopping
      • TerminateOnNaN
      • AllPredictionsSame
      • EmailNotification
      • LogUpload
      • DataSubsetTestRun
      • FunctionOnTrainLoop
    • ddp
      • DistributedSamplerSetEpoch
      • InMultiProcessDataLoad
    • gradient
      • GradientCallbackBase
      • GradValueClip
      • GradNormClip
      • GradientStatsPrint
      • GradDistributionPlot
    • model_load
      • ModelLoadContinueTraining
    • model_save
      • ModelCheckpoint
      • ModelIterationCheckpoint
      • ModelTrainEndSave
    • performance_eval
      • ModelPerformanceEvaluation
      • ModelPerformancePrintReport
      • TrainHistoryFormatter
      • MetricHistoryRename
      • ModelTrainHistoryBaseCB
      • ModelTrainHistoryPlot
      • ModelTrainHistoryFileWriter
    • tensorboard
      • TensorboardReporterBaseCB
      • TensorboardTrainBatchLoss
      • TensorboardTrainHistoryMetric
      • TensorboardFullTracking
    • train_schedule
    • wandb
      • AlertConfig
      • WandBTracking
• data
  • Submodules
    • batch_model_feed_defs
      • AbstractModelFeedDefinition
    • dataset
      • BasicDataset
      • ListDataset
• schedulers
  • Submodules
    • basic
      • AbstractScheduler
      • GeneralLRSchedulerCallback
      • ReduceLROnPlateauScheduler
      • ReduceLROnPlateauMetricScheduler
      • LambdaLRScheduler
      • StepLRScheduler
      • MultiStepLRScheduler
    • warmup
      • ConstantWithWarmupScheduler
      • CosineWithWarmupScheduler
      • HardRestartsCosineWithWarmupScheduler
      • LinearWithWarmupScheduler
• train_loop
  • Subpackages
    • components
      • Submodules
  • Submodules
    • train_loop
      • TrainLoop
    • train_loop_tracking
      • TrainLoopCheckpoint
      • TrainLoopEndSave
      • TrainLoopCheckpointEndSave

Submodules

• model
  • TTModel
    • TTModel.get_loss()
    • TTModel.get_loss_eval()
    • TTModel.get_predictions()
    • TTModel.training
  • TTBasicModel
    • TTBasicModel.get_loss()
    • TTBasicModel.get_predictions()
    • TTBasicModel.training
  • TTBasicMultiGPUModel
    • TTBasicMultiGPUModel.get_loss()
    • TTBasicMultiGPUModel.training
  • MultiGPUModelWrap
    • MultiGPUModelWrap.forward()
    • MultiGPUModelWrap.training
  • ModelWrap
• model_predict
  • PyTorchModelPredictor
    • PyTorchModelPredictor.model_predict()
    • PyTorchModelPredictor.model_get_loss()
    • PyTorchModelPredictor.evaluate_model()
    • PyTorchModelPredictor.evaluate_result_package()
    • PyTorchModelPredictor.execute_batch_end_callbacks()
    • PyTorchModelPredictor.execute_epoch_end_callbacks()
    • PyTorchModelPredictor.evaluate_metric()
    • PyTorchModelPredictor.evaluate_metric_list()
• multi_loss_optim
  • MultiLoss
    • MultiLoss.backward()
    • MultiLoss.item()
    • MultiLoss.numpy()
    • MultiLoss.detach()
    • MultiLoss.cpu()
    • MultiLoss.cuda()
    • MultiLoss.to()
    • MultiLoss.device
    • MultiLoss.get_loss_dict()
  • MultiOptimizer
    • MultiOptimizer.step()
    • MultiOptimizer.zero_grad()
    • MultiOptimizer.state_dict()
    • MultiOptimizer.load_state_dict()
• parallel
  • TTParallelBase
    • TTParallelBase.get_loss()
    • TTParallelBase.get_loss_eval()
    • TTParallelBase.get_predictions()
  • TTDataParallel
    • TTDataParallel.training
  • TTDistributedDataParallel
    • TTDistributedDataParallel.training