refactor(recipes): drop redundant entrypoint override from glm53-flash-nvfp4

The day-0 vLLM image's default ENTRYPOINT is already [vllm serve], so the
explicit override was redundant; let the image supply it (model + args are
appended to the entrypoint as-is).

recipes/glm53-flash-nvfp4/recipe.toml

@@ -55,8 +55,6 @@ pipeline_parallel_size = 1
 distributed_backend = "mp"
 shm_size = "32gb"
 
-entrypoint = ["vllm", "serve"]
-
 args = [
     # Stable served name — quant-agnostic (shared across glm53 quants)
     "--served-model-name", "glm53",