mindspore_lite

The Python API only support cloud-side inference.

Context

mindspore_lite.Context

The Context class is used to transfer environment variables during execution.

Converter

mindspore_lite.FmkType	When Converter, the FmkType is used to define Input model framework type.
mindspore_lite.Converter	Constructs a Converter class.

Model

mindspore_lite.Model	The Model class defines a MindSpore Lite's model, facilitating computational graph management.
mindspore_lite.ModelGroup	The ModelGroup class is used to define a MindSpore model group, facilitating multiple models to share workspace memory or weights (including constants and variables) memory or both.
mindspore_lite.ModelGroupFlag	The ModelGroupFlag class defines the type of the model group.
mindspore_lite.ModelParallelRunner	The ModelParallelRunner class defines a MindSpore Lite's Runner, which support model parallelism.
mindspore_lite.ModelType	The ModelType class defines the type of the model exported or imported in MindSpot Lite.

Tensor

mindspore_lite.DataType	The DataType class defines the data type of the Tensor in MindSpore Lite.
mindspore_lite.Format	The Format class defines the format of the Tensor in MindSpore Lite.
mindspore_lite.Tensor	The Tensor class defines a Tensor in MindSpore Lite.
mindspore_lite.TensorMeta	The TensorMeta class defines a TensorInfo in MindSpore Lite.

LLMEngine

mindspore_lite.LLMReq	LLMEngine request, used to represent a multi round inference task.
mindspore_lite.LLMEngineStatus	LLMEngine Status, which can be got from LLEngine.fetch_status.
mindspore_lite.LLMRole	Role of LLMEngine.
mindspore_lite.LLMEngine	The LLMEngine class defines a MindSpore Lite's LLMEngine, used to load and manage Large Language Mode, and schedule and execute inference request.
mindspore_lite.LLMStatusCode	LLM Error Code
mindspore_lite.LLMException	Base Error class for LLM

LiteInfer

mindspore_lite.LiteInfer(model_or_net, ...)

The LiteInfer class takes training model as input and performs predictions directly.