引導式測驗 · 中等 等級

NVIDIA NIM 推論微服務測驗

評估自託管 AI 推理服務的 NIM 容器、API 整合、硬體需求和部署檢查。

相關引導路徑Nvidia Nim 推論微服務
問題 1 的 8

What does an NVIDIA NIM container primarily provide?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. What does an NVIDIA NIM container primarily provide?
    • A packaged inference service for supported models
    • A replacement for the host GPU driver
    • A guarantee of model accuracy
    • An automatic data-labeling workflow
  2. Why check the supported model and hardware requirements before deployment?
    • Every container runs every model identically
    • Support and memory needs vary by NIM, model and GPU
    • API compatibility guarantees sufficient GPU capacity
    • The host driver is included in all images
  3. What should an application verify about a NIM API?
    • That every provider has identical semantics
    • Only that the URL uses HTTP
    • Specific request fields, streaming and error behavior for that NIM
    • That model licenses need no review
  4. Which metric helps assess an LLM NIM serving workload?
    • The count of package folders
    • Only the image file size
    • Training accuracy on old data
    • Time to first token and throughput under representative prompts
  5. Why keep client application logic separated from a NIM endpoint?
    • It guarantees all models have the same outputs
    • It makes it easier to change serving backends or versions
    • It removes authentication needs
    • It prevents GPU memory use
  6. What remains the host's responsibility when running a GPU container?
    • The model's response schema only
    • The container's model license only
    • Compatible driver and GPU runtime configuration
    • The app's prompt template
  7. Why review model and container licensing before deployment?
    • A license replaces security scanning
    • NIM containers remove all software licenses
    • Licensing determines model latency
    • Usage conditions may restrict how artifacts can be used or redistributed
  8. What does a NIM container not provide automatically?
    • Governance, evaluation and access-control decisions
    • A serving endpoint
    • A packaged runtime for supported use
    • An API described in its documentation