SageMaker Qwen Factual Validation Results Quiz
Multiple Choice Questions
- What did the September 2 recheck find?
- A) Zero remaining resources
- B) One Unified Studio project still
ACTIVE - C) One running EKS GPU cluster
- D) A completed Training Job
Show Answer
Answer: B
The historical record reported reclaimed App, S3, and IAM resources with one project remaining. The September 12 documentation review did not re-query current AWS state.
- What was the key failure in attempt 3?
- A) A misspelled model ID
- B) Missing project membership
- C) A dataset hash mismatch
- D) CUDA out of memory
Show Answer
Answer: B
The caller role group profile was not assigned as an owner/member.
- Which results were not published?
- A) Record count
- B) MLflow App version
- C) Fine-tuned F1 and GPU cost
- D) Python test count
Show Answer
Answer: C
GPU training was not executed, so those comparison values were not measured. This does not imply zero total cost for resources such as the MLflow App and S3.
- What is the first rerun gate?
- A) Submit the full Job
- B) Reconcile inventory and resolve remaining or unknown experiment-owned resources
- C) Create an EKS cluster
- D) Publish tuned metrics
Show Answer
Answer: B
Permission errors or absence from a list do not prove deletion. A shared name prefix does not authorize deleting someone else's resources.
- How do the target architecture and actual workflow differ?
- A) Both prove completion
- B) One is a rerun design; the other records the real stop point
- C) Both are cost reports
- D) The actual workflow shows completed GPU training
Show Answer
Answer: B
The historical workflow stops before GPU training. It does not prove the current account state or zero total experiment cost.