NVIDIA-Certified Associate: Generative AI Multimodal 試験対策問題集

模擬試験モードの設定

上記は本サイトの模擬試験モードの設定値です。実際の試験の出題数・制限時間・合格基準は、各認定プログラムの公式サイトでご確認ください。

出題分野

• Experimentation（実験と評価）（出題比率 25%）
• Core Machine Learning and AI Knowledge（機械学習とAIの基礎）（出題比率 20%）
• Multimodal Data（マルチモーダルデータ）（出題比率 15%）
• Software Development（ソフトウェア開発）（出題比率 15%）
• Data Analysis and Visualization（データ分析と可視化）（出題比率 11%）
• Performance Optimization（性能の最適化）（出題比率 9%）
• Trustworthy AI（信頼できるAI）（出題比率 5%）

フリー問題

1. A dataset of product photographs was collected by scraping a catalog. Many produ…
2. A team must report how close their image generator is to real photographs. A rev…
3. A speech dataset is assembled from three sources: call-center recordings at 8 kH…
4. A team is starting work on a product that must answer questions about photograph…
5. A caption-generation system was changed in three ways at once: the image encoder…
6. In a training batch of image-and-caption pairs, a contrastive objective computes…
7. A deep network is built by stacking many blocks. Each block adds its own transfo…
8. A team trains a model that takes audio and video together. The audio branch conv…
9. A team trains a model that takes an image and a sentence and learns to place mat…
10. A team stores embeddings for two million photographs and must find the closest m…
11. A retrieval system indexes photograph embeddings. The team decides to switch to …
12. A team must sort incoming product photographs into 40 categories that the mercha…
13. A team wants to let users supply a reference photograph so generated images matc…
14. A generation service calls a model that occasionally returns an output that fail…
15. A team uses a contrastively trained image-and-text model in two places: to condi…
16. A team must show how a speech model's error rate varies across 12 recording cond…
17. A team reviews a collection of 500,000 image-and-caption pairs before training. …
18. A team adapts a large pretrained image-and-text model to a narrow inspection tas…
19. A team has budget for 40 training runs to tune five hyperparameters. A grid with…
20. A team publishes a document alongside a released multimodal model describing its…