No Feedback When Generating Random Data

[robertgshaw2-redhat]

Is your feature request related to a problem? Please describe.

• When generating a long prompt like the below:

guidellm benchmark --target=$URL --model=$MODEL --rate-type=concurrent --rate=50 --max-requests=50 --output-path=~/guidellm_results.json --processor=$MODEL --data='{"prompt_tokens":8000, "output_tokens":5000}'

We get stuck generating the random data for about 1 minute:

(app-root) bash-5.1$ guidellm benchmark --target=$URL --model=$MODEL --rate-type=concurrent --rate=50 --max-requests=50 --output-path=~/guidellm_results.json --processor=$MODEL --data='{"prompt_tokens":8000, "output_tokens":5000}'
Creating backend...
Backend openai_http connected to http://10.16.1.195:8000 for model                    
meta-llama/Meta-Llama-3-8B-Instruct.                                                  
Creating request loader...

Describe the solution you'd like
There should be some feedback during this period about what is happening. I thought there was a bug in vllm or in guidellm during this period

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

[markurtz]

Related to #270