[docs] Benchmark Serving Incorrect Arg (vllm-project#25474)

Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
Signed-off-by: charlifu <charlifu@amd.com>

Author: vllmellm

docs/contributing/benchmarks.md

@@ -680,7 +680,7 @@ vllm bench serve \
   --save-result \
   --result-dir ~/vllm_benchmark_results \
   --save-detailed \
-  --endpoint /v1/chat/completion
+  --endpoint /v1/chat/completions
 ```
 
 ##### Videos (ShareGPT4Video)
@@ -707,7 +707,7 @@ vllm bench serve \
   --save-result \
   --result-dir ~/vllm_benchmark_results \
   --save-detailed \
-  --endpoint /v1/chat/completion
+  --endpoint /v1/chat/completions
 ```
 
 ##### Synthetic Random Images (random-mm)