mindspore

Data Presentation

Tensor

`mindspore.Tensor`	Tensor is a data structure that stores an n-dimensional array.
`mindspore.tensor`	Create a new Tensor in Cell.construct() or function decorated by @jit.
`mindspore.COOTensor`	A sparse representation of a set of nonzero elements from a tensor at given indices.
`mindspore.CSRTensor`	Constructs a sparse tensor in CSR (Compressed Sparse Row) format, with specified values indicated by values and row and column positions indicated by indptr and indices.
`mindspore.RowTensor`	A sparse representation of a set of tensor slices at given indices.
`mindspore.SparseTensor`	A sparse representation of a set of nonzero elements from a tensor at given indices.

Parameter

`mindspore.Parameter`	Parameter is a Tensor subclass, when they are assigned as Cell attributes they are automatically added to the list of its parameters, and will appear, e.g.
`mindspore.ParameterTuple`	Inherited from tuple, ParameterTuple is used to save multiple parameter.

DataType

class mindspore.dtype

Create a data type object of MindSpore.

The actual path of dtype is /mindspore/common/dtype.py. Run the following command to import the package:

from mindspore import dtype as mstype

Numeric Type

Currently, MindSpore supports int type, uint type, float type and complex type. The following table lists the details.

Definition	Description
`mindspore.int8` , `mindspore.byte`	8-bit integer
`mindspore.int16` , `mindspore.short`	16-bit integer
`mindspore.int32` , `mindspore.intc`	32-bit integer
`mindspore.int64` , `mindspore.intp`	64-bit integer
`mindspore.uint8` , `mindspore.ubyte`	unsigned 8-bit integer
`mindspore.uint16` , `mindspore.ushort`	unsigned 16-bit integer
`mindspore.uint32` , `mindspore.uintc`	unsigned 32-bit integer
`mindspore.uint64` , `mindspore.uintp`	unsigned 64-bit integer
`mindspore.float16` , `mindspore.half`	16-bit floating-point number
`mindspore.float32` , `mindspore.single`	32-bit floating-point number
`mindspore.float64` , `mindspore.double`	64-bit floating-point number
`mindspore.bfloat16`	16-bit brain-floating-point number
`mindspore.complex64`	64-bit complex number
`mindspore.complex128`	128-bit complex number

Other Type

For other defined types, see the following table.

Type	Description
`tensor`	MindSpore's `tensor` type. Data format uses NCHW. For details, see tensor.
`bool_`	Boolean `True` or `False`.
`int_`	Integer scalar.
`uint`	Unsigned integer scalar.
`float_`	Floating-point scalar.
`complex`	Complex scalar.
`number`	Number, including `int_` , `uint` , `float_` , `complex` and `bool_` .
`list_`	List constructed by `tensor` , such as `List[T0,T1,...,Tn]` , where the element `Ti` can be of different types.
`tuple_`	Tuple constructed by `tensor` , such as `Tuple[T0,T1,...,Tn]` , where the element `Ti` can be of different types.
`function`	Function. Return in two ways, when function is not None, returns Func directly, the other returns Func(args: List[T0,T1,…,Tn], retval: T) when function is None.
`type_type`	Type definition of type.
`type_none`	No matching return type, corresponding to the `type(None)` in Python.
`symbolic_key`	The value of a variable is used as a key of the variable in `env_type` .
`env_type`	Used to store the gradient of the free variable of a function, where the key is the `symbolic_key` of the free variable's node and the value is the gradient.

class mindspore.common.np_dtype

np_dtype expands Numpy's data types.

The actual path of np_dtype is /mindspore/common/np_dtype.py. Run the following command to import the package:

from mindspore.common import np_dtype

Numeric Type

Type	Description
`bfloat16`	The `bfloat16` data type under NumPy. This type is only used to construct Tensor of type `bfloat16`, and does not guarantee the full computing power under Numpy. Takes effect only if the version of Numpy at runtime is not less than the version of Numpy at compilation.

`mindspore.dtype_to_nptype`	Convert MindSpore dtype to numpy data type.
`mindspore.dtype_to_pytype`	Convert MindSpore dtype to python data type.
`mindspore.pytype_to_dtype`	Convert python type to MindSpore type.
`mindspore.get_py_obj_dtype`	Get the MindSpore data type, which corresponds to python type or variable.
`mindspore.QuantDtype`	An enum for quant datatype, contains INT1 ~ INT16, UINT1 ~ UINT16.

Context

`mindspore.set_context`	Set context for running environment.
`mindspore.get_context`	Get context attribute value according to the input key.
`mindspore.set_auto_parallel_context`	Set auto parallel context, only data parallel supported on CPU.
`mindspore.get_auto_parallel_context`	Get auto parallel context attribute value according to the key.
`mindspore.reset_auto_parallel_context`	Reset auto parallel context attributes to the default values.
`mindspore.ParallelMode`	Parallel mode options.
`mindspore.set_ps_context`	Set parameter server training mode context.
`mindspore.get_ps_context`	Get parameter server training mode context attribute value according to the key.
`mindspore.reset_ps_context`	Reset parameter server training mode context attributes to the default values.
`mindspore.set_algo_parameters`	Set parameters in the algorithm for parallel strategy searching.
`mindspore.get_algo_parameters`	Get the algorithm parameter config attributes.
`mindspore.reset_algo_parameters`	Reset the algorithm parameter attributes.
`mindspore.set_offload_context`	Configure heterogeneous training detailed parameters to adjust the offload strategy.
`mindspore.get_offload_context`	Gets the offload configuration parameters.

Seed

`mindspore.set_seed`	Set global seed.
`mindspore.get_seed`	Get global seed.

Random Number Generator

`mindspore.get_rng_state`	Get the state of the default generator.
`mindspore.Generator`	A generator that manages the state of random numbers and provides seed and offset for random functions.
`mindspore.initial_seed`	Return the initial seed of the default generator.
`mindspore.manual_seed`	Set the default generator seed.
`mindspore.seed`	Seed the default generator with random number.
`mindspore.set_rng_state`	Set the state of the default generator.

Serialization

`mindspore.async_ckpt_thread_status`	Get the status of asynchronous save checkpoint thread.
`mindspore.build_searched_strategy`	Build strategy of every parameter in network.
`mindspore.check_checkpoint`	Check whether the checkpoint is valid.
`mindspore.convert_model`	Convert mindir model to other format model.
`mindspore.export`	Export the MindSpore network into an offline model in the specified format.
`mindspore.load`	Load MindIR.
`mindspore.load_checkpoint`	Load checkpoint info from a specified file.
`mindspore.load_checkpoint_async`	Load checkpoint info from a specified file asyncly.
`mindspore.load_distributed_checkpoint`	Load checkpoint into net for distributed predication.
`mindspore.load_mindir`	load protobuf file.
`mindspore.load_param_into_net`	Load parameters into network, return parameter list that are not loaded in the network.
`mindspore.load_segmented_checkpoints`	Load checkpoint info from a specified file.
`mindspore.merge_pipeline_strategys`	Merge parallel strategy between all pipeline stages in pipeline parallel mode.
`mindspore.merge_sliced_parameter`	Merge parameter slices into one parameter.
`mindspore.obfuscate_model`	Obfuscate a model of MindIR format.
`mindspore.parse_print`	Parse data file generated by `mindspore.ops.Print`.
`mindspore.rank_list_for_transform`	List of original distributed checkpoint rank index for obtaining the target checkpoint of a rank_id during the distributed checkpoint conversion.
`mindspore.restore_group_info_list`	Build rank list, the checkpoint of ranks in the rank list has the same contents with the local rank who saves the group_info_file_name.
`mindspore.save_checkpoint`	Save checkpoint to a specified file.
`mindspore.save_mindir`	save protobuf file.
`mindspore.transform_checkpoint_by_rank`	Transform distributed checkpoint from source sharding strategy to destination sharding strategy by rank for a network.
`mindspore.transform_checkpoints`	Transform distributed checkpoint from source sharding strategy to destination sharding strategy for a rank.

Automatic Differentiation

`mindspore.grad`	A wrapper function to generate the gradient function for the input function.
`mindspore.value_and_grad`	A wrapper function to generate the function to calculate forward output and gradient for the input function.
`mindspore.get_grad`	When return_ids of `mindspore.grad()` or `mindspore.grad()` is set to `True` , use return value of mindspore.grad, or the second return value of mindspore.grad as gradients.
`mindspore.jacfwd`	Compute Jacobian via forward mode, corresponding to forward-mode differentiation.
`mindspore.jacrev`	Compute Jacobian via reverse mode, corresponding to reverse-mode differentiation.
`mindspore.jvp`	Compute the jacobian-vector-product of the given network.
`mindspore.vjp`	Compute the vector-jacobian-product of the given network.

Parallel Optimization

Automatic Vectorization

mindspore.vmap

Vectorizing map (vmap) is a kind of higher-order function to map fn along the parameter axes.

Parallel

`mindspore.Layout`	Parallel layout describes the detailed sharding information.
`mindspore.parameter_broadcast`	Broadcast parameter to other rank in data parallel dimension.
`mindspore.recompute`	This function is used to reduce memory, when run block, rather than storing the intermediate activation computed in forward pass, we will recompute it in backward pass.
`mindspore.reshard`	Specify the tensor by the given layout.
`mindspore.shard`	Defining the input and output layouts of this cell and the parallel strategies of remaining ops will be generated by sharding propagation.
`mindspore.sync_pipeline_shared_parameters`	synchronize pipeline parallel stage shared parameters.

JIT

`mindspore.JitConfig`	Jit config for compile.
`mindspore.jit`	Create a callable MindSpore graph from a Python function.
`mindspore.jit_class`	Class decorator for user-defined classes.
`mindspore.ms_class`	Class decorator for user-defined classes.
`mindspore.ms_function`	Create a callable MindSpore graph from a Python function.
`mindspore.ms_memory_recycle`	Recycle memory used by MindSpore.
`mindspore.mutable`	Make a constant value mutable.
`mindspore.constexpr`	Used to calculate constant in graph copmpiling process and improve compile performance in GRAPH_MODE.
`mindspore.lazy_inline`	Make the cell to be reusable.
`mindspore.no_inline`	Make the function to be reusable.

Tool

Dataset Helper

`mindspore.DatasetHelper`	DatasetHelper is a class to process the MindData dataset and provides the information of dataset.
`mindspore.Symbol`	Symbol is a data structure to indicate the symbolic info of shape.
`mindspore.connect_network_with_dataset`	Connect the network with dataset in dataset_helper.
`mindspore.data_sink`	A wrapper function to generate a function for the input function.

Debugging and Tuning

`mindspore.Profiler`	This class to enable the profiling of MindSpore neural networks.
`mindspore.SummaryCollector`	SummaryCollector can help you to collect some common information, such as loss, learning late, computational graph and so on.
`mindspore.SummaryLandscape`	SummaryLandscape can help you to collect loss landscape information.
`mindspore.SummaryRecord`	SummaryRecord is used to record the summary data and lineage data.
`mindspore.set_dump`	Enable or disable dump for the target and its contents.

Log

`mindspore.get_level`	Get the logger level.
`mindspore.get_log_config`	Get logger configurations.

Installation Verification

mindspore.run_check

Provide a convenient API to check if the installation is successful or failed.