deepmd.tf.train.trainer

Attributes

log

Classes

DPTrainer
DatasetLoader	Generate an OP that loads the training data from the given DeepmdDataSystem.

Functions

_is_subdir(→ bool)

Module Contents

deepmd.tf.train.trainer.log

deepmd.tf.train.trainer._is_subdir(path: str, directory: str) → bool

class deepmd.tf.train.trainer.DPTrainer(jdata: dict, run_opt: Any, is_compress: bool = False)

run_opt

is_compress = False

_init_param(jdata: dict) → None

build(data: deepmd.tf.utils.data_system.DeepmdDataSystem | None = None, stop_batch: int = 0, origin_type_map: list[str] | None = None, suffix: str = '') → None

_build_lr() → None

_build_loss() → tuple[None, None] | tuple[deepmd.tf.env.tf.Tensor, dict[str, deepmd.tf.env.tf.Tensor]]

_build_network(data: deepmd.tf.utils.data_system.DeepmdDataSystem, suffix: str = '') → None

_build_optimizer() → Any

_build_training() → None

_init_session() → None

train(train_data: dict | None = None, valid_data: dict | None = None) → None

save_checkpoint(cur_batch: int) → None

get_feed_dict(batch: dict, is_training: bool) → dict

get_global_step() → int

valid_on_the_fly(fp: TextIO, train_batches: list, valid_batches: list, print_header: bool = False, fitting_key: str | None = None) → None

static print_header(fp: TextIO, train_results: dict, valid_results: dict | None) → None

static print_on_training(fp: TextIO, train_results: dict, valid_results: dict | None, cur_batch: int, cur_lr: float) → None

static eval_single_list(single_batch_list: list | None, loss: deepmd.tf.loss.loss.Loss, sess: deepmd.tf.env.tf.Session, get_feed_dict_func: collections.abc.Callable, prefix: str = '') → dict | None

get_evaluation_results(batch_list: list) → dict

save_compressed() → None

Save the compressed graph.

_get_place_holders(data_dict: dict) → None

_init_from_frz_model() → None

_init_from_ckpt(ckpt_meta: str) → None

_init_from_pretrained_model(data: deepmd.tf.utils.data_system.DeepmdDataSystem, origin_type_map: list[str] | None = None, bias_adjust_mode: str = 'change-by-statistic') → None

Init the embedding net variables with the given frozen model.

Parameters:

dataDeepmdDataSystem

The training data.

origin_type_maplist

The original type_map in dataset, they are targets to change the energy bias.

bias_adjust_modestr

The mode for changing energy bias : [‘change-by-statistic’, ‘set-by-statistic’] ‘change-by-statistic’ : perform predictions on energies of target dataset,

and do least square on the errors to obtain the target shift as bias.

‘set-by-statistic’ : directly use the statistic energy bias in the target dataset.

_change_energy_bias(data: deepmd.tf.utils.data_system.DeepmdDataSystem, frozen_model: str, origin_type_map: list[str] | None, bias_adjust_mode: str = 'change-by-statistic') → None

property data_requirements: list[deepmd.utils.data.DataRequirementItem]

class deepmd.tf.train.trainer.DatasetLoader(train_data: deepmd.tf.utils.data_system.DeepmdDataSystem)

Generate an OP that loads the training data from the given DeepmdDataSystem.

It can be used to load the training data in the training process, so there is no waiting time between training steps.

Parameters:

train_dataDeepmdDataSystem

The training data.

Examples

>>> loader = DatasetLoader(train_data)
>>> data_op = loader.build()
>>> with tf.Session() as sess:
>>>     data_list = sess.run(data_op)
>>> data_dict = loader.get_data_dict(data_list)

train_data

data_keys

data_types

build() → list[deepmd.tf.env.tf.Tensor]

Build the OP that loads the training data.

Returns:

list[tf.Tensor]

Tensor of the loaded data.

get_data_dict(batch_list: list[numpy.ndarray]) → dict[str, numpy.ndarray]

Generate a dict of the loaded data.

Parameters:

batch_listlist[np.ndarray]

The loaded data.

Returns:

dict[str, np.ndarray]

The dict of the loaded data.