fix(cli): Add validation for search-mode parameter

[mayani-nv]

This minor PR will add validation check for the search-mode parameter. Originally i tried to run the cuvs_bench as following

python -m cuvs_bench.run --dataset fbv6_50m --algorithms cuvs_cagra --batch-size 100 -k 10
Enter the path to dataset folder [/opt/conda/lib/python3.10/site-packages/cuvs_bench/run/datasets/]: /data/benchmarks/cuvs_experiments/datasets
Enter the comma separated groups of parameters [base]: base
Enter the search mode ("latency" or "throughput") [latency]: througput
/opt/conda/lib/python3.10/site-packages/cuvs_bench/run/run.py:187: UserWarning: Could not load YAML config

As you can see, I had the typo in this line Enter the search mode ("latency" or "throughput") [latency]: througput missing h and it defaulted to latency. This PR would print error if there's anything else specified other than latency or throughput.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[divyegala]

/ok to test 4432dc8

[divyegala]

/ok to test 4be381c

[cjnolet]

/ok to test 7f6ddb3

[cjnolet]

/merge