aclmdlBundleInitFromMem

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product / Atlas A3 inference product	√
Atlas A2 training product / Atlas A2 inference product	√
Atlas 200I/500 A2 inference product	√
Atlas inference product	√
Atlas training product	√

Initializes a model from the memory if variables need to be dynamically updated during model execution.

      
           aclError aclmdlBundleInitFromMem(const void* model, size_t modelSize, void *varWeightPtr, size_t varWeightSize, uint32_t *bundleId)

Parameter	Input/Output	Description
model	Input	Pointer to the memory address for storing model data. The memory is managed by the user and cannot be freed during aclmdlBundleLoadModel loading. The model file is built using the graph build API. Call aclgrphBundleBuildModel to compile the model, call aclgrphBundleSaveModel to save the model, and then load the saved .om model file into the memory. For details about the graph APIs, see aclgrphBundleBuildModel in Graph Development. In Ascend EP form, allocate memory on the host. In Ascend RC form, allocate memory on the device. In Ctrl CPU open form, allocate memory on the device.
modelSize	Input	Size of the model data in the memory, in bytes.
varWeightPtr	Input	Pointer to the address of the refreshable weight memory required by the model. The memory is managed by the user and cannot be freed during model execution. If a null pointer is passed, the memory is managed by the system.
varWeightSize	Input	Size of the refreshable weight memory required by the model, in bytes. Invalid when varWeightPtr is a null pointer.
bundleId	Output	Bundle ID returned after the bundle model is successfully initialized. The bundle ID is used for model identification in subsequent operations.

0 on success; else, failure. For details, see aclError.

For details about the API call sequence and sample code, see Weight update (user-managed memory).

Parent topic: Model Loading and Unloading