Online Access Free NCA-GENM Exam Questions

You are deploying a multimodal Generative A1 model on a cloud platform. The model takes video and text as input to generate video descriptions. The model's performance needs to be monitored to ensure it meets certain performance SLAs. Which of the following metrics are MOST crucial to monitor in a production environment to ensure both computational efficiency and output quality? (Select TWO)

• A.GPU utilization.
• B.Inference latency (time per request).
• C.Number of lines of code in the model.
• D.BLEU score (or similar text generation metric) for generated descriptions.
• E.Model size on disk.

You are building a Generative Adversarial Network (GAN) to generate high-resolution images. The generated images suffer from mode collapse, where the generator only produces a limited variety of images. Which of the following techniques would be MOST effective in mitigating mode collapse?

• A.Increasing the learning rate of the generator.
• B.Decreasing the learning rate of the discriminator.
• C.Using a simpler generator architecture.
• D.Reducing the size of the generator's latent space.
• E.Using mini-batch discrimination or feature matching in the discriminator.

You are tasked with generating realistic images of human faces using a GAN. However, you notice that the generated images often contain artifacts, such as distorted facial features or unrealistic textures. Which of the following techniques would be most effective in improving the realism and quality of the generated faces?

• A.Applying L1 regularization to the generator's weights.
• B.Training the GAN for fewer epochs.
• C.Using a smaller batch size.
• D.Using a simpler discriminator architecture.
• E.Employing a StyleGAN architecture with adaptive instance normalization (AdalN) and mapping network.

You are developing a multimodal generative model that takes a text description as input and generates a corresponding image. However, you notice that the generated images often lack fine-grained details and realism. Which of the following approaches could you employ to improve the quality and realism of the generated images? (Select all that apply)

• A.Implement a loss function that encourages the generated images to match the statistical distribution of real images.
• B.Train the model using a generative adversarial network (GAN) framework.
• C.Use a higher-resolution image generator architecture.
• D.Use a smaller training dataset.
• E.Decrease the size of the text encoder.

You are tasked with building a system that can generate captions for images. You want to use a transformer-based model. During inference, you notice that the model tends to generate repetitive captions. Which of the following decoding strategies could you use to mitigate this issue?

• A.Greedy decoding
• B.Random sampling
• C.Beam search with a high beam width
• D.Top-k sampling
• E.Beam search with a length penalty