mindspore.train
mindspore.train.summary
SummaryRecord.
User can use SummaryRecord to dump the summary data, the summary is a series of operations to collect data for analysis and visualization.
- class mindspore.train.summary.SummaryRecord(log_dir, queue_max_size=0, flush_time=120, file_prefix='events', file_suffix='_MS', network=None, max_file_size=None)[source]
SummaryRecord is used to record the summary data and lineage data.
The API will create a summary file and lineage files lazily in a given directory and writes data to them. It writes the data to files by executing the ‘record’ method. In addition to recording the data bubbled up from the network by defining the summary operators, SummaryRecord also supports to record extra data which can be added by calling add_value.
Note
Make sure to close the SummaryRecord at the end, otherwise the process will not exit. Please see the Example section below to learn how to close properly in two ways.
Only one SummaryRecord instance is allowed at a time, otherwise it will cause data writing problems.
- Parameters
log_dir (str) – The log_dir is a directory location to save the summary.
queue_max_size (int) – Deprecated. The capacity of event queue.(reserved). Default: 0.
flush_time (int) – Deprecated. Frequency of flush the summary file to disk. The unit is second. Default: 120.
file_prefix (str) – The prefix of file. Default: “events”.
file_suffix (str) – The suffix of file. Default: “_MS”.
network (Cell) – Obtain a pipeline through network for saving graph summary. Default: None.
max_file_size (Optional[int]) – The maximum size of each file that can be written to disk (in bytes). Unlimited by default. For example, to write not larger than 4GB, specify max_file_size=4 * 1024**3.
- Raises
TypeError – If the data type of max_file_size, queue_max_size or flush_time is not int, or the data type of file_prefix and file_suffix is not str.
RuntimeError – If the log_dir is not a normalized absolute path name.
Examples
>>> # use in with statement to auto close >>> with SummaryRecord(log_dir="./summary_dir") as summary_record: >>> pass >>> >>> # use in try .. finally .. to ensure closing >>> try: >>> summary_record = SummaryRecord(log_dir="./summary_dir") >>> finally: >>> summary_record.close()
- add_value(plugin, name, value)[source]
Add value to be recorded later.
When the plugin is ‘tensor’, ‘scalar’, ‘image’ or ‘histogram’, the name should be the tag name, and the value should be a Tensor.
When the plugin is ‘graph’, the value should be a GraphProto.
When the plugin is ‘dataset_graph’, ‘train_lineage’, ‘eval_lineage’, or ‘custom_lineage_data’, the value should be a proto message.
- Parameters
plugin (str) – The value of the plugin.
name (str) – The value of the name.
value (Union[Tensor, GraphProto, TrainLineage, EvaluationLineage, DatasetGraph, UserDefinedInfo]) –
The value to store.
The data type of value should be ‘GraphProto’ when the plugin is ‘graph’.
The data type of value should be ‘Tensor’ when the plugin is ‘scalar’, ‘image’, ‘tensor’ or ‘histogram’.
The data type of value should be ‘TrainLineage’ when the plugin is ‘train_lineage’.
The data type of value should be ‘EvaluationLineage’ when the plugin is ‘eval_lineage’.
The data type of value should be ‘DatasetGraph’ when the plugin is ‘dataset_graph’.
The data type of value should be ‘UserDefinedInfo’ when the plugin is ‘custom_lineage_data’.
- Raises
ValueError – When the name is not valid.
TypeError – When the value is not a Tensor.
Examples
>>> with SummaryRecord(log_dir="./summary_dir", file_prefix="xxx_", file_suffix="_yyy") as summary_record: >>> summary_record.add_value('scalar', 'loss', Tensor(0.1))
- close()[source]
Flush all events and close summary records. Please use the statement to autoclose.
Examples
>>> try: >>> summary_record = SummaryRecord(log_dir="./summary_dir") >>> finally: >>> summary_record.close()
- flush()[source]
Flush the event file to disk.
Call it to make sure that all pending events have been written to disk.
Examples
>>> with SummaryRecord(log_dir="./summary_dir", file_prefix="xxx_", file_suffix="_yyy") as summary_record: >>> summary_record.flush()
- property log_dir
Get the full path of the log file.
- Returns
str, the full path of log file.
Examples
>>> with SummaryRecord(log_dir="./summary_dir", file_prefix="xxx_", file_suffix="_yyy") as summary_record: >>> print(summary_record.log_dir)
- record(step, train_network=None, plugin_filter=None)[source]
Record the summary.
- Parameters
- Returns
bool, whether the record process is successful or not.
Examples
>>> with SummaryRecord(log_dir="./summary_dir", file_prefix="xxx_", file_suffix="_yyy") as summary_record: >>> summary_record.record(step=2)
- set_mode(mode)[source]
Set the mode for the recorder to be aware. The mode is set to ‘train’ by default.
- Parameters
mode (str) – The mode to be set, which should be ‘train’ or ‘eval’.
- Raises
ValueError – When the mode is not recognized.
Examples
>>> with SummaryRecord(log_dir="./summary_dir", file_prefix="xxx_", file_suffix="_yyy") as summary_record: >>> summary_record.set_mode('eval')
mindspore.train.callback
Callback related classes and functions.
- class mindspore.train.callback.Callback[source]
Abstract base class used to build a callback class. Callbacks are context managers which will be entered and exited when passing into the Model. You can use this mechanism to initialize and release resources automatically.
Callback function will execute some operations in the current step or epoch.
Examples
>>> class Print_info(Callback): >>> def step_end(self, run_context): >>> cb_params = run_context.original_args() >>> print(cb_params.cur_epoch_num) >>> print(cb_params.cur_step_num) >>> >>> print_cb = Print_info() >>> model.train(epoch, dataset, callbacks=print_cb)
- begin(run_context)[source]
Called once before the network executing.
- Parameters
run_context (RunContext) – Include some information of the model.
- end(run_context)[source]
Called once after network training.
- Parameters
run_context (RunContext) – Include some information of the model.
- epoch_begin(run_context)[source]
Called before each epoch beginning.
- Parameters
run_context (RunContext) – Include some information of the model.
- epoch_end(run_context)[source]
Called after each epoch finished.
- Parameters
run_context (RunContext) – Include some information of the model.
- step_begin(run_context)[source]
Called before each epoch beginning.
- Parameters
run_context (RunContext) – Include some information of the model.
- step_end(run_context)[source]
Called after each step finished.
- Parameters
run_context (RunContext) – Include some information of the model.
- class mindspore.train.callback.CheckpointConfig(save_checkpoint_steps=1, save_checkpoint_seconds=0, keep_checkpoint_max=5, keep_checkpoint_per_n_minutes=0, integrated_save=True, async_save=False)[source]
The configuration of model checkpoint.
Note
During the training process, if dataset is transmitted through the data channel, It is suggested to set ‘save_checkpoint_steps’ to an integer multiple of loop_size. Otherwise, the time to save the checkpoint may be biased.
- Parameters
save_checkpoint_steps (int) – Steps to save checkpoint. Default: 1.
save_checkpoint_seconds (int) – Seconds to save checkpoint. Default: 0. Can’t be used with save_checkpoint_steps at the same time.
keep_checkpoint_max (int) – Maximum number of checkpoint files can be saved. Default: 5.
keep_checkpoint_per_n_minutes (int) – Keep one checkpoint every n minutes. Default: 0. Can’t be used with keep_checkpoint_max at the same time.
integrated_save (bool) – Whether to perform integrated save function in automatic model parallel scene. Default: True. Integrated save function is only supported in automatic parallel scene, not supported in manual parallel.
async_save (bool) – Whether asynchronous execution saves the checkpoint to a file. Default: False
- Raises
ValueError – If the input_param is None or 0.
Examples
>>> config = CheckpointConfig() >>> ckpoint_cb = ModelCheckpoint(prefix="ck_prefix", directory='./', config=config) >>> model.train(10, dataset, callbacks=ckpoint_cb)
- property async_save
Get the value of _async_save.
- property integrated_save
Get the value of _integrated_save.
- property keep_checkpoint_max
Get the value of _keep_checkpoint_max.
- property keep_checkpoint_per_n_minutes
Get the value of _keep_checkpoint_per_n_minutes.
- property save_checkpoint_seconds
Get the value of _save_checkpoint_seconds.
- property save_checkpoint_steps
Get the value of _save_checkpoint_steps.
- class mindspore.train.callback.LossMonitor(per_print_times=1)[source]
Monitor the loss in training.
If the loss is NAN or INF, it will terminate training.
Note
If per_print_times is 0, do not print loss.
- Parameters
per_print_times (int) – Print the loss each every time. Default: 1.
- Raises
ValueError – If print_step is not an integer or less than zero.
- class mindspore.train.callback.ModelCheckpoint(prefix='CKP', directory=None, config=None)[source]
The checkpoint callback class.
It is called to combine with train process and save the model and network parameters after traning.
- Parameters
prefix (str) – The prefix name of checkpoint files. Default: “CKP”.
directory (str) – The path of the folder which will be saved in the checkpoint file. Default: None.
config (CheckpointConfig) – Checkpoint strategy configuration. Default: None.
- Raises
ValueError – If the prefix is invalid.
TypeError – If the config is not CheckpointConfig type.
- end(run_context)[source]
Save the last checkpoint after training finished.
- Parameters
run_context (RunContext) – Context of the train running.
- property latest_ckpt_file_name
Return the latest checkpoint path and file name.
- step_end(run_context)[source]
Save the checkpoint at the end of step.
- Parameters
run_context (RunContext) – Context of the train running.
- class mindspore.train.callback.RunContext(original_args)[source]
Provides information about the model.
Provides information about original request to model function. Callback objects can stop the loop by calling request_stop() of run_context.
- Parameters
original_args (dict) – Holding the related information of model.
- get_stop_requested()[source]
Returns whether a stop is requested or not.
- Returns
bool, if true, model.train() stops iterations.
- class mindspore.train.callback.SummaryCollector(summary_dir, collect_freq=10, collect_specified_data=None, keep_default_action=True, custom_lineage_data=None, collect_tensor_freq=None, max_file_size=None)[source]
SummaryCollector can help you to collect some common information.
It can help you to collect loss, learning late, computational graph and so on. SummaryCollector also enables the summary operator to collect data from a summary file.
Note
Multiple SummaryCollector instances in callback list are not allowed.
Not all information is collected at the training phase or at the eval phase.
SummaryCollector always record the data collected by the summary operator.
- Parameters
summary_dir (str) – The collected data will be persisted to this directory. If the directory does not exist, it will be created automatically.
collect_freq (int) – Set the frequency of data collection, it should be greater then zero, and the unit is step. Default: 10. If a frequency is set, we will collect data when (current steps % freq) equals to 0, and the first step will be collected at any time. It is important to note that if the data sink mode is used, the unit will become the epoch. It is not recommended to collect data too frequently, which can affect performance.
collect_specified_data (Union[None, dict]) –
Perform custom operations on the collected data. Default: None. By default, if set to None, all data is collected as the default behavior. You can customize the collected data with a dictionary. For example, you can set {‘collect_metric’: False} to control not collecting metrics. The data that supports control is shown below.
collect_metric: Whether to collect training metrics, currently only the loss is collected. The first output will be treated as the loss and it will be averaged. Optional: True/False. Default: True.
collect_graph: Whether to collect the computational graph. Currently, only training computational graph is collected. Optional: True/False. Default: True.
collect_train_lineage: Whether to collect lineage data for the training phase, this field will be displayed on the lineage page of Mindinsight. Optional: True/False. Default: True.
collect_eval_lineage: Whether to collect lineage data for the evaluation phase, this field will be displayed on the lineage page of Mindinsight. Optional: True/False. Default: True.
collect_input_data: Whether to collect dataset for each training. Currently only image data is supported. Optional: True/False. Default: True.
collect_dataset_graph: Whether to collect dataset graph for the training phase. Optional: True/False. Default: True.
histogram_regular: Collect weight and bias for parameter distribution page and displayed in MindInsight. This field allows regular strings to control which parameters to collect. Default: None, it means only the first five parameters are collected. It is not recommended to collect too many parameters at once, as it can affect performance. Note that if you collect too many parameters and run out of memory, the training will fail.
keep_default_action (bool) – This field affects the collection behavior of the ‘collect_specified_data’ field. Optional: True/False, Default: True. True: it means that after specified data is set, non-specified data is collected as the default behavior. False: it means that after specified data is set, only the specified data is collected, and the others are not collected.
custom_lineage_data (Union[dict, None]) – Allows you to customize the data and present it on the MingInsight lineage page. In the custom data, the type of the key supports str, and the type of value supports str, int and float. Default: None, it means there is no custom data.
collect_tensor_freq (Optional[int]) – The same semantics as the collect_freq, but controls TensorSummary only. Because TensorSummary data is too large to be compared with other summary data, this parameter is used to reduce its collection. By default, The maximum number of steps for collecting TensorSummary data is 20, but it will not exceed the number of steps for collecting other summary data. Default: None, which means to follow the behavior as described above. For example, given collect_freq=10, when the total steps is 600, TensorSummary will be collected 20 steps, while other summary data 61 steps, but when the total steps is 20, both TensorSummary and other summary will be collected 3 steps. Also note that when in parallel mode, the total steps will be splitted evenly, which will affect the number of steps TensorSummary will be collected.
max_file_size (Optional[int]) – The maximum size in bytes of each file that can be written to the disk. Default: None, which means no limit. For example, to write not larger than 4GB, specify max_file_size=4 * 1024**3.
- Raises
ValueError – If the parameter value is not expected.
TypeError – If the parameter type is not expected.
RuntimeError – If an error occurs during data collection.
Examples
>>> # Simple usage: >>> summary_collector = SummaryCollector(summary_dir='./summary_dir') >>> model.train(epoch, dataset, callbacks=summary_collector) >>> >>> # Do not collect metric and collect the first layer parameter, others are collected by default >>> specified={'collect_metric': False, 'histogram_regular': '^conv1.*'} >>> summary_collector = SummaryCollector(summary_dir='./summary_dir', collect_specified_data=specified) >>> model.train(epoch, dataset, callbacks=summary_collector) >>> >>> # Only collect metric, custom lineage data and record data that collected by the summary operator, >>> # others are not collected >>> specified = {'collect_metric': True} >>> summary_collector = SummaryCollector('./summary_dir', >>> collect_specified_data=specified, >>> keep_default_action=False, >>> custom_lineage_data={'version': 'resnet50_v1'} >>> ) >>> model.train(epoch, dataset, callbacks=summary_collector)
mindspore.train.serialization
Model and parameters serialization.
- mindspore.train.serialization.build_searched_strategy(strategy_filename)[source]
Build strategy of every parameter in network.
- Parameters
strategy_filename (str) – Name of strategy file.
- Returns
Dictionary, whose key is parameter name and value is slice strategy of this parameter.
- Raises
ValueError – Strategy file is incorrect.
TypeError – Strategy_filename is not str.
Examples
>>> strategy_filename = "./strategy_train.ckpt" >>> strategy = build_searched_strategy(strategy_filename)
- mindspore.train.serialization.export(net, *inputs, file_name, file_format='AIR')[source]
Export the MindSpore prediction model to a file in the specified format.
- Parameters
net (Cell) – MindSpore network.
inputs (Tensor) – Inputs of the net.
file_name (str) – File name of the model to be exported.
file_format (str) –
MindSpore currently supports ‘AIR’, ‘ONNX’ and ‘MINDIR’ format for exported model.
AIR: Ascend Intermidiate Representation. An intermidiate representation format of Ascend model. Recommended suffix for output file is ‘.air’.
ONNX: Open Neural Network eXchange. An open format built to represent machine learning models. Recommended suffix for output file is ‘.onnx’.
MINDIR: MindSpore Native Intermidiate Representation for Anf. An intermidiate representation format for MindSpore models. Recommended suffix for output file is ‘.mindir’.
- mindspore.train.serialization.load_checkpoint(ckpt_file_name, net=None)[source]
Loads checkpoint info from a specified file.
- Parameters
- Returns
Dict, key is parameter name, value is a Parameter.
- Raises
ValueError – Checkpoint file is incorrect.
- mindspore.train.serialization.load_param_into_net(net, parameter_dict)[source]
Loads parameters into network.
- mindspore.train.serialization.merge_sliced_parameter(sliced_parameters, strategy=None)[source]
Merge parameter slices to one whole parameter.
- Parameters
sliced_parameters (list[Parameter]) – Parameter slices in order of rank_id.
strategy (dict) –
Parameter slice strategy, the default is None. If strategy is None, just merge parameter slices in 0 axis order.
key (str): Parameter name.
value (<class ‘node_strategy_pb2.ParallelLayouts’>): Slice strategy of this parameter.
- Returns
Parameter, the merged parameter which has the whole data.
- Raises
ValueError – Failed to merge.
TypeError – The sliced_parameters is incorrect or strategy is not dict.
KeyError – The parameter name is not in keys of strategy.
Examples
>>> strategy = build_searched_strategy("./strategy_train.ckpt") >>> sliced_parameters = [ >>> Parameter(Tensor(np.array([0.00023915, 0.00013939, -0.00098059])), >>> "network.embedding_table"), >>> Parameter(Tensor(np.array([0.00015815, 0.00015458, -0.00012125])), >>> "network.embedding_table"), >>> Parameter(Tensor(np.array([0.00042165, 0.00029692, -0.00007941])), >>> "network.embedding_table"), >>> Parameter(Tensor(np.array([0.00084451, 0.00089960, -0.00010431])), >>> "network.embedding_table")] >>> merged_parameter = merge_sliced_parameter(sliced_parameters, strategy)
- mindspore.train.serialization.parse_print(print_file_name)[source]
Loads Print data from a specified file.
- Parameters
print_file_name (str) – The file name of saved print data.
- Returns
List, element of list is Tensor.
- Raises
ValueError – The print file may be empty, please make sure enter the correct file name.
- mindspore.train.serialization.save_checkpoint(parameter_list, ckpt_file_name, async_save=False)[source]
Saves checkpoint info to a specified file.
- Parameters
- Raises
RuntimeError – Failed to save the Checkpoint file.
mindspore.train.amp
Auto mixed precision.
- mindspore.train.amp.build_train_network(network, optimizer, loss_fn=None, level='O0', **kwargs)[source]
Build the mixed precision training cell automatically.
- Parameters
network (Cell) – Definition of the network.
loss_fn (Union[None, Cell]) – Definition of the loss_fn. If None, the network should have the loss inside. Default: None.
optimizer (Optimizer) – Optimizer to update the Parameter.
level (str) –
Supports [O0, O2, O3]. Default: “O0”.
O0: Do not change.
O2: Cast network to float16, keep batchnorm and loss_fn (if set) run in float32, using dynamic loss scale.
O3: Cast network to float16, with additional property ‘keep_batchnorm_fp32=False’.
O2 is recommended on GPU, O3 is recommended on Ascend.
cast_model_type (
mindspore.dtype) – Supports mstype.float16 or mstype.float32. If set to mstype.float16, use float16 mode to train. If set, overwrite the level setting.keep_batchnorm_fp32 (bool) – Keep Batchnorm run in float32. If set, overwrite the level setting. Only cast_model_type is float16, keep_batchnorm_fp32 will take effect.
loss_scale_manager (Union[None, LossScaleManager]) – If None, not scale the loss, or else scale the loss by LossScaleManager. If set, overwrite the level setting.
mindspore.train.loss_scale_manager
Loss scale manager abstract class.
- class mindspore.train.loss_scale_manager.DynamicLossScaleManager(init_loss_scale=16777216, scale_factor=2, scale_window=2000)[source]
Dynamic loss-scale manager.
- Parameters
Examples
>>> loss_scale_manager = DynamicLossScaleManager() >>> model = Model(net, loss_scale_manager=loss_scale_manager)
- class mindspore.train.loss_scale_manager.FixedLossScaleManager(loss_scale=128.0, drop_overflow_update=True)[source]
Fixed loss-scale manager.
- Parameters
Examples
>>> loss_scale_manager = FixedLossScaleManager() >>> model = Model(net, loss_scale_manager=loss_scale_manager)
mindspore.train.quant
Quantization.
User can use quantization aware to train a model. MindSpore supports quantization aware training, which models quantization errors in both the forward and backward passes using fake-quantization operations. Note that the entire computation is carried out in floating point. At the end of quantization aware training, MindSpore provides conversion functions to convert the trained model into lower precision.
- mindspore.train.quant.convert_quant_network(network, bn_fold=True, freeze_bn=10000000, quant_delay=(0, 0), num_bits=(8, 8), per_channel=(False, False), symmetric=(False, False), narrow_range=(False, False))[source]
Create quantization aware training network.
- Parameters
network (Cell) – Obtain a pipeline through network for saving graph summary.
bn_fold (bool) – Flag to used bn fold ops for simulation inference operation. Default: True.
freeze_bn (int) – Number of steps after which BatchNorm OP parameters used total mean and variance. Default: 1e7.
quant_delay (int, list or tuple) – Number of steps after which weights and activations are quantized during eval. The first element represent weights and second element represent data flow. Default: (0, 0)
num_bits (int, list or tuple) – Number of bits to use for quantize weights and activations. The first element represent weights and second element represent data flow. Default: (8, 8)
per_channel (bool, list or tuple) – Quantization granularity based on layer or on channel. If True then base on per channel otherwise base on per layer. The first element represent weights and second element represent data flow. Default: (False, False)
symmetric (bool, list or tuple) – Whether the quantization algorithm is symmetric or not. If True then base on symmetric otherwise base on asymmetric. The first element represent weights and second element represent data flow. Default: (False, False)
narrow_range (bool, list or tuple) – Whether the quantization algorithm uses narrow range or not. The first element represents weights and the second element represents data flow. Default: (False, False)
- Returns
Cell, Network which has change to quantization aware training network cell.