NVIDIA Deep Learning Institute: Fundamentals of Deep Learning Practice Test

An engineer exports a trained ResNet-50 to ONNX, then uses the TensorRT Python API on an A10 GPU to build an engine with FP16 enabled and INT8 calibration using a calibration dataset of 800 representative images. After building, the engine achieves 3.8x higher throughput than the original PyTorch FP32 model at batch size 8, with top-1 accuracy dropping only 0.6 points. Which statement correctly explains both the speedup and the small accuracy change?

An NLP engineer trains a unidirectional LSTM language model on 50-token sequences. After convergence, the team adds a second LSTM layer on top of the first (stacked) and also wraps both with dropout=0.25 between time steps and between layers. They observe that the validation perplexity improves by 4 points relative to the single-layer baseline, but training perplexity is now higher than validation perplexity for the first 15 epochs. What is the most accurate interpretation?

An ML engineer is implementing a CNN for 32x32 RGB CIFAR-10 images. The first layer is a 3x3 convolution with 32 filters, stride=1, padding=1, followed by ReLU and 2x2 max-pooling with stride=2. When the input resolution is doubled to 64x64 while keeping every other layer identical, what is the spatial size of the feature map immediately after the first pooling layer?

A researcher trains a two-hidden-layer MLP on the Iris dataset using stochastic gradient descent with a learning rate of 0.01. The model reaches 98% training accuracy but the decision boundaries appear overly smooth and fail to capture the non-linear separation between two of the classes in a held-out visualization. Which modification to the network architecture would most directly increase its capacity to model more complex decision boundaries?

A developer builds two TensorRT engines from the identical ONNX graph of a U-Net segmentation model. Engine A uses FP16 with no INT8. Engine B additionally enables INT8 with a calibration dataset of 1200 images and sets the builder flag for INT8. On an Orin AGX, Engine B runs 1.9x faster than Engine A but produces visibly blocky artifacts on fine structures (thin lines, small text). Which statement best explains the artifacts?