[BE] Update vllm-benchmark parameter descriptions (#47)

* [no ci] Update parameter description to support release and pull request checks

Signed-off-by: Huy Do <huydhn@gmail.com>

* Minor tweak

Signed-off-by: Huy Do <huydhn@gmail.com>

* [no ci] Another tweak

Signed-off-by: Huy Do <huydhn@gmail.com>

---------

Signed-off-by: Huy Do <huydhn@gmail.com>

Author: huydhn

.github/workflows/vllm-benchmark.yml

@@ -7,22 +7,22 @@ on:
   workflow_dispatch:
     inputs:
       vllm_branch:
-        description: vLLM branch
+        description: vLLM branch (main, releases/vERSION for release validation, or refs/pull/PR_NUMBER/head for pre-merge check on pull request)
         required: true
         type: string
         default: main
       vllm_commit:
-        description: vLLM commit
+        description: vLLM commit (optional, default to the latest commit in the branch that has not yet been benchmarked)
         required: false
         type: string
       models:
         description: |
-          A comma-separated list of models to benchmark, leave empty to run everything
+          A comma-separated list of models from vllm-benchmarks/benchmarks (optional, default to run everything)
         required: false
         type: string
       runners:
         description: |
-          A comma-separated list of runners to run the benchmark, i.e. h100, mi300, spr, emr
+          A comma-separated list of runners from .github/scripts/generate_vllm_benchmark_matrix.py to run the benchmark (optional, default to run everything)
         required: true
         type: string
         default: h100,mi300,spr