train_loop

class aitoolbox.torchtrain.train_loop.train_loop.TrainLoop(model, train_loader, validation_loader, test_loader, optimizer, criterion, collate_batch_pred_fn=<function append_predictions>, pred_transform_fn=<function torch_cat_transf>, end_auto_eval=True, lazy_experiment_save=False, print_callbacks=False, gpu_mode='single', cuda_device_idx=None, use_amp=False)[source]

Bases: object

Core PyTorch TrainLoop supporting the model training and target prediction

Implements core training procedures: batch feeding into the network as part of (multi)epoch train loop, calculation of the loss & gradients. Apart from training related functionality the TrainLoop also implements the logic needed for prediction of target variables.

Parameters:

  • model (TTModel or ModelWrap or TTDataParallel) – neural network model

  • train_loader (torch.utils.data.DataLoader) – data loader for train data set

  • validation_loader (torch.utils.data.DataLoader or None) – data loader for validation data set

  • test_loader (torch.utils.data.DataLoader or None) – data loader for test data set

  • optimizer (torch.optim.Optimizer or MultiOptimizer) – optimizer algorithm.

  • criterion (torch.nn.Module or MultiLoss or None) – criterion during the training procedure

  • collate_batch_pred_fn (callable) – collate function transforming batch predictions as they come out from the model

  • pred_transform_fn (callable) – function transforming all the produced predictions after all the batches have been run through the model

  • end_auto_eval (bool or int) – used to optionally disable otherwise automatic end of epoch/training val/test loss calculations. This is useful when conducting very costly experiments to save on compute time. Specify either True/False boolean to always run or never run after each epoch or specify an int to execute only every specified number of epochs.

  • lazy_experiment_save (bool) – when in lazy mode experiment tracking components will create the experiment folder only after some training results are available (possibly at the end of the first epoch) instead of at the beginning of training.

  • print_callbacks (bool) – at the start of training print the list of registered callbacks which will be executed during the run of the train loop

  • gpu_mode (str) –

    GPU training mode selection. TrainLoop supports different GPU training modes by specifying one of the following:

    • 'single': single GPU training

    • 'dp': multi-GPU training via DataParallel

    • 'ddp': multi-GPU training via DistributedDataParallel

  • cuda_device_idx (int or None) – CUDA device index used when training on multiple GPUs

  • use_amp (bool or dict) –

    Use 16-bit Automatic Mixed Precision (AMP).

    To switch to AMP mode either:

    • set this parameter to True to use default AMP GradScaler initialization params

    • provide custom AMP GradScaler initialization parameters as a dict as this parameter

fit(num_epochs=0, num_iterations=0, callbacks=None, grad_accumulation=1, **kwargs)[source]

Train the model using the train loop

This is the general API method which starts the model training. By calling this method and depending on the selected training mode provided as the TrainLoop’s gpu_mode parameter the training will start in one of the following training modes:

  • Basic (CPU or single GPU) mode

  • DataParallel mode

  • DistributedDataParallel mode

Parameters:

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list or None) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

  • **kwargs –

    additional parameters for training methods:

    These training methods are called by the TrainLoop depending on the specified setting of the TrainLoop’s gpu_mode parameter.

Returns:

trained model

Return type:

TTModel or torch.nn.Module or TTDataParallel

_train(num_epochs, num_iterations, callbacks=None, grad_accumulation=1)[source]

Train the model using the train loop

Parameters:

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list or None) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

Returns:

trained model

Return type:

TTModel or torch.nn.Module or TTDataParallel

_calculate_batch_loss(batch_data)[source]

Push batch data through the model and calculate the batch loss

Parameters:

batch_data (torch.Tensor) – input data batch

Returns:

loss calculated on current batch

Return type:

loss (torch.Tensor or MultiLoss)

_backward_pass(loss_batch, optimizer_idx)[source]

Execute backward pass from the current batch loss

Parameters:

  • loss_batch (torch.Tensor or MultiLoss) – loss calculated on current batch

  • optimizer_idx (int) – index of the current optimizer. Mostly useful when using multiple optimizers. When only a single optimizer is used this parameter can be ignored.

Returns:

None

_optimizer_step(optimizer_idx)[source]

Execute the optimizer step

Parameters:

optimizer_idx (int) – index of the current optimizer. Mostly useful when using multiple optimizers. When only a single optimizer is used this parameter can be ignored.

Returns:

None

_optimizer_zero_grad(optimizer_idx)[source]

Execute optimizer zero grad

Parameters:

optimizer_idx (int) – index of the current optimizer. Mostly useful when using multiple optimizers. When only a single optimizer is used this parameter can be ignored.

Returns:

None

should_execute_optimizer_update()[source]

Determine if optimizer update based on calculated gradients should be done at the current iteration

Combined with optimizer update we normally also execute zero_grad as well as different gradient clipping operations.

This method is especially important in the case when gradient accumulation is used in training. It provides knowledge when model parameter updates via the optimizer are made based on accumulated gradients.

Note

Switched from a simple condition to better a condition to also cover the final non-complete batch:

if (self.iteration + 1) % self.grad_accumulation == 0

if (self.iteration + 1) % self.grad_accumulation == 0 or self.iteration == len(self.train_loader) - 1

Returns:

if in current iteration a model parameter update via the optimizer should be done

Return type:

bool

auto_execute_end_of_epoch()[source]

Basic performance evaluation executed by default at the end of each epoch

Mainly evaluation of the loss functions which are always present as part of the training loop.

Returns:

None

auto_execute_end_of_training()[source]

Basic performance evaluation executed by default at the end of the training process

Returns:

None

parse_loss(loss_record)[source]

Helper function to process different possible loss formats

Primarily useful for parsing between single loss representation and the multi-loss representation.

Parameters:

loss_record (list) –

list of Tensor losses from each processed batch.

If we used single loss than the loss_record is a list of Tensors where each element Tensor is loss for a single batch.

If we used multiple losses wrapped inside MultiLoss(), these behave the same way as normal dicts as MultiLoss subclasses a dict and thus implements dict protocols. Consequently, loss_record can be thought as a list of (MultiLoss) dicts, where each dict represents a loss for a single batch:

[MultiLoss({'loss_1': Tensor(1.), 'loss_2': Tensor(33.)}), MultiLoss({ ... })]

Returns:

in the case of single loss torch Tensor is returned, otherwise the dict of multiple losses is returned where each value is again a torch Tensor

Note

Important to note: all the returned loss Tensors are left on the original device (e.g. a GPU).

Return type:

torch.DoubleTensor or MultiLoss

_print_save_loss(loss_parsed, loss_type_name, loss_print_description)[source]

Helper function which prints information about parsed loss and saves the loss results into the history

Parameters:

  • loss_parsed (torch.Tensor or MultiLoss) – parsed loss result either as a single value or as MultiLoss in case of multiple losses

  • loss_type_name (str) – type of the provided loss result

  • loss_print_description (str) – presentation description text of the provided loss result

Returns:

None

evaluate_loss_on_train_set(force_prediction=False, float_dict_format=False)[source]

Run train dataset through the network without updating the weights and return the loss

Parameters:

  • force_prediction (bool) – recompute the loss even if it is available in the prediction cache. This causes the old cached value to be overwritten.

  • float_dict_format (bool) – if true, simplified loss representation is returned. In case of single loss, a float is returned, while in case of multi-loss a dict extracted from MultiLoss wrapper is returned. If false, the standard torch.Tensor or MultiLoss get returned.

Returns:

train set loss. Returned tensors are on the CPU. Depending on the set float_dict_format parameter either a standard or simplified loss representation is returned: torch.Tensor/MultiLoss vs. float/dict

Return type:

torch.Tensor or MultiLoss or float or dict

evaluate_loss_on_validation_set(force_prediction=False, float_dict_format=False)[source]

Run validation dataset through the network without updating the weights and return the loss

Parameters:

  • force_prediction (bool) – recompute the loss even if it is available in the prediction cache. This causes the old cached value to be overwritten.

  • float_dict_format (bool) – if true, simplified loss representation is returned. In case of single loss, a float is returned, while in case of multi-loss a dict extracted from MultiLoss wrapper is returned. If false, the standard torch.Tensor or MultiLoss get returned.

Returns:

validation set loss. Returned tensors are on the CPU. Depending on the set float_dict_format parameter either a standard or simplified loss representation is returned: torch.Tensor/MultiLoss vs. float/dict

Return type:

torch.Tensor or MultiLoss or float or dict

evaluate_loss_on_test_set(force_prediction=False, float_dict_format=False)[source]

Run test dataset through the network without updating the weights and return the loss

Parameters:

  • force_prediction (bool) – recompute the loss even if it is available in the prediction cache. This causes the old cached value to be overwritten.

  • float_dict_format (bool) – if true, simplified loss representation is returned. In case of single loss, a float is returned, while in case of multi-loss a dict extracted from MultiLoss wrapper is returned. If false, the standard torch.Tensor or MultiLoss get returned.

Returns:

test set loss. Returned tensors are on the CPU. Depending on the set float_dict_format parameter either a standard or simplified loss representation is returned: torch.Tensor/MultiLoss vs. float/dict

Return type:

torch.Tensor or MultiLoss or float or dict

evaluate_model_loss(data_loader, move_to_cpu=False, dataset_info=None)[source]

Run given dataset through the network without updating the weights and return the loss

Parameters:
Returns:

Calculated average loss over all the batches. In the case of multi loss, the MultiLoss wrapper gets returned.

Note

Important to note: by default the returned loss tensors are left on the same device as they are computed. Meaning, that the returned values can potentially still be on the GPU.

Return type:

torch.Tensor or MultiLoss

predict_on_train_set(force_prediction=False, execute_callbacks=False)[source]

Run train dataset through the network and return true target values, target predictions and metadata

Parameters:

  • force_prediction (bool) – recompute the output prediction even if it is available in the prediction cache. This causes the old cached predictions to be overwritten.

  • execute_callbacks (bool) – If true, prediction loop will execute provided callbacks after prediction for each batch has been made. Otherwise, callbacks at this position are ignored.

Returns:

y_pred, y_true, metadata in the form of dict of lists/torch.Tensors/np.arrays

Return type:

(torch.Tensor, torch.Tensor, dict)

predict_on_validation_set(force_prediction=False, execute_callbacks=False)[source]

Run validation dataset through the network and return true target values, target predictions and metadata

Parameters:

  • force_prediction (bool) – recompute the output prediction even if it is available in the prediction cache. This causes the old cached predictions to be overwritten.

  • execute_callbacks (bool) – If true, prediction loop will execute provided callbacks after prediction for each batch has been made. Otherwise, callbacks at this position are ignored.

Returns:

y_pred, y_true, metadata in the form of dict of lists/torch.Tensors/np.arrays

Return type:

(torch.Tensor, torch.Tensor, dict)

predict_on_test_set(force_prediction=False, execute_callbacks=False)[source]

Run test dataset through the network and return true target values, target predictions and metadata

Parameters:

  • force_prediction (bool) – recompute the output prediction even if it is available in the prediction cache. This causes the old cached predictions to be overwritten.

  • execute_callbacks (bool) – If true, prediction loop will execute provided callbacks after prediction for each batch has been made. Otherwise, callbacks at this position are ignored.

Returns:

y_pred, y_true, metadata in the form of dict of lists/torch.Tensors/np.arrays

Return type:

(torch.Tensor, torch.Tensor, dict)

predict_with_model(data_loader, execute_callbacks=False, move_to_cpu=False, dataset_info=None)[source]

Run given dataset through the network and return true target values, target predictions and metadata

Parameters:

  • data_loader (torch.utils.data.DataLoader) – dataloader containing the data on which the output predictions are calculated

  • execute_callbacks (bool) – If true, prediction loop will execute provided callbacks after prediction for each batch has been made. Otherwise, callbacks at this position are ignored.

  • move_to_cpu (bool) – should the predicted returned results be moved to the CPU. Otherwise, the returned results are kept on the original device (which can also be a GPU).

  • dataset_info (dict or None) – additional information describing the dataset inside the provided dataloader. One such dataset info is the dataset type (train, validation, or test) set by predict_on_train_set(), predict_on_validation_set() and predict_on_test_set() methods.

Returns:

y_pred, y_true, metadata in the form of dict of lists/torch.Tensors/np.arrays

Return type:

(torch.Tensor, torch.Tensor, dict)

insert_metric_result_into_history(metric_name, metric_result)[source]

Insert a metric result into the train history

This is the main and preferred API function for metric insertion as part of the train loop.

Parameters:

  • metric_name (str) – name of the metric to be inserted

  • metric_result (float or dict) – new result for the corresponding metric

get_schedulers()[source]

Get the registered schedulers

Schedulers in TrainLoop training are implemented as callbacks under the hood.

Returns:

list of scheduler (callbacks)

Return type:

list

get_num_training_steps()[source]

Get the number of actual training steps

Useful in case of gradient accumulation to learn the number of steps where the gradient is actually updated in between the accumulation steps.

Returns:

number of training steps / iterations

Return type:

int

is_main_process()[source]

Is current process the main training process

In case of single GPU/CPU we have single process so this function is always True. However, for DDP training main process is treated as that which is at rank 0.

Returns:

if current process is the main training process. In case of DDP it is process at rank 0

Return type:

bool

static convert_loss_to_float_dict_format(loss)[source]

Util method for converting loss records in Tensor/MultiLoss format into simpler float/dict format

Parameters:

loss (torch.Tensor or MultiLoss) – more complex loss representation. In case of single loss it is torch Tensor. In case of multi-loss it is MultiLoss wrapper.

Returns:

simplified loss representation. In case of single loss it is a single float value. In case of multi-loss it is a dict extracted out from the given MultiLoss wrapper.

Return type:

float or dict

_train_dp(num_epochs, num_iterations, callbacks=None, grad_accumulation=1, dp_model_args=None)[source]

Train the model on multi-GPU with DataParallel auto wrapping

Parameters:

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list or None) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

  • dp_model_args (dict or None) – parameters for aitoolbox.torchtrain.parallel.TTDataParallel / torch.nn.DataParallel DP model wrap.

Returns:

trained model

Return type:

TTDataParallel or torch.nn.DataParallel

_train_ddp(num_epochs, num_iterations, callbacks=None, grad_accumulation=1, ddp_model_args=None, in_process_data_load=None, num_nodes=1, node_rank=0, num_gpus=0, backend='nccl', init_method='env://', on_gpu=True)[source]

Train the model using the train loop in the Distributed Data Parallel setting

During the training, multiple processes will be spawned, one for each of the available GPUs.

Parameters:

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list or None) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

  • ddp_model_args (dict or None) – parameters for underlying PyTorch DistributedDataParallel model

  • in_process_data_load (AbstractCallback or list or None) – in-process data loading logic implemented as a torchtrain callback. The logic should be placed inside the on_multiprocess_start() callback function. When using this data loading option bear in mind that loaded dataset will be replicated in memory for every spawned training process. This can in turn in cause extensive overall memory consumption.

  • num_nodes (int) – number of nodes in the cluster

  • node_rank (int) – rank of the current node

  • num_gpus (int) – number of GPUs in the node

  • backend (str) – The backend to use. For more information look up the documentation for torch.distributed.init_process_group(). Valid values include mpi, gloo, and nccl.

  • init_method (str) – URL specifying how to initialize the process group. For more information look up the documentation for torch.distributed.init_process_group().

  • on_gpu (bool) – if the DDP training is executed on the GPU or on the CPU

_spawn_fit(gpu, ddp_args, num_epochs, num_iterations, callbacks, grad_accumulation, in_process_data_load)[source]

Helper function that prepares the TrainLoop state inside each of the spawned processes and initiates training

Parameters:

  • gpu (int) – provided by the mp.spawn(); index of the GPU allocated to the current process

  • ddp_args (dict) – parameters dict needed for the distributed training setup

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list or None) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

  • in_process_data_load (list or None) – in-process data loading logic implemented as a torchtrain callback. The logic should be placed inside the on_multiprocess_start() callback function. When using this data loading option bear in mind that loaded dataset will be replicated in memory for every spawned training process. This can in turn in cause extensive overall memory consumption.

__call__(num_epochs=0, num_iterations=0, callbacks=None, grad_accumulation=1, **kwargs)[source]

Train the model using the train loop

This is a convenience function which calls the main TrainLoop model training method fit().

Parameters:

  • num_epochs (int) – how many epochs the network will be trained

  • num_iterations (int) – how many iterations (batches) the network will be trained. This enables more granular specification of the training length than the num_epochs parameter.

  • callbacks (list) – callbacks that are executed during the training run

  • grad_accumulation (int) – number of batches the gradients are accumulated before updating weights

  • **kwargs – additional parameters for _train_dp() and _train_ddp() methods.

Returns:

trained model

Return type:

TTModel or torch.nn.Module or TTDataParallel