Free HCIA-AI V3.5 Exam H13-311_V3.5 Exam Practice Test

Huawei Cloud ModelArts provides ModelBox for device-edge-cloud joint development. Which of the following are its optimization policies?

• Hardware affinity
• Operator optimization
• Automatic segmentation of operators
• Model replication

Which of the following statements about datasets are true?

• Testing refers to a process that uses a trained model for prediction. The dataset, which is used for testing, is called a testing set, and each sample is called a test sample.
• A dataset generally has multiple dimensions. In each dimension, events or attributes that reflect the performance or nature of a sample in a particular aspect are called features.
• In machine learning, a dataset is generally divided into a training set, validation set, and test set.
• When it comes to the machine learning process, the validation set and the test set are essentially the same.

When you use MindSpore to execute the following code, which of the following is the output?from mindspore import ops import mindsporeshape = (2, 2) ones = ops.Ones()output = ones(shape, dtype=mindspore.float32) print(output)

• [[1 1]
• [1 1]]
• [[1. 1.]
• [1. 1.]]
• 1
• [[1. 1.
• 1. 1.]]

AI inference chips need to be optimized and are thus more complex than those used for training.

• TRUE
• FALSE

Which of the following are common gradient descent methods?

• Batch gradient descent (BGD)
• Mini-batch gradient descent (MBGD)
• Multi-dimensional gradient descent (MDGD)
• Stochastic gradient descent (SGD)

Which of the following is NOT a key feature that enables all-scenario deployment and collaboration for MindSpore?

• Data and computing graphs are transmitted to Ascend AI Processors.
• Federal meta-learning enables real-time, coordinated model updates between different devices, and across the device and cloud.
• Unified model IR delivers a consistent deployment experience.
• Graph optimization based on a software-hardware synergy shields the differences between scenarios.

Which of the following statements is false about gradient descent algorithms?

• Each time the global gradient updates its weight, all training samples need to be calculated.
• When GPUs are used for parallel computing, the mini-batch gradient descent (MBGD) takes less time than the stochastic gradient descent (SGD) to complete an epoch.
• The global gradient descent is relatively stable, which helps the model converge to the global extremum.
• When there are too many samples and GPUs are not used for parallel computing, the convergence process of the global gradient algorithm is time-consuming.

Which of the following is NOT a commonly used AI computing framework?

• PyTorch
• MindSpore
• TensorFlow
• OpenCV

Huawei's full-stack AI solution includes Ascend, MindSpore, and ModelArts. (Enter an acronym.)

• AII
• AIIS
• CANN
• None of the above

HarmonyOS can provide AI capabilities for external systems only through the integrated HMS Core.

• TRUE
• FALSE