Fix benchmark script for saving benchmark result

[lsy323]

if args.additional_metadata_metrics_to_save is None, json.loads would fail. args.additional_metadata_metrics_to_save is None by default. Adding a check to avoid the failure.

Test:
Using the benchmark example script in README

python JetStream/benchmarks/benchmark_serving.py \ --tokenizer ~/maxtext/assets/tokenizer.llama2 \ --warmup-mode full \ --save-result \ --save-request-outputs \ --request-outputs-file-path outputs.json \ --num-prompts 1000 \ --max-output-length 1024 \ --dataset openorca 

[FanhaiLu1]

Thanks for fixing it!