A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 16:00 UTC
[Misc] add description attribute in CLI (#15921)
Signed-off-by: reidliu41 <reid201711@gmail.com> Co-authored-by: reidliu41 <reid201711@gmail.com>
Reid committed
Apr 8, 2025 at 06:30 UTC
fad6e2538ebacbed4ba9cb04f0d143b7daf2be29
4 files changed
+9
-3
vllm/entrypoints/cli/benchmark/base.py
+1
@@ -32,6 +32,7 @@ class BenchmarkSubcommandBase(CLISubcommand):
32
parser = subparsers.add_parser(
33
self.name,
34
help=self.help,
35
+ description=self.help,
36
usage=f"vllm bench {self.name} [options]")
37
self.add_cli_args(parser)
38
return parser
vllm/entrypoints/cli/benchmark/main.py
+1
@@ -33,6 +33,7 @@ class BenchmarkSubcommand(CLISubcommand):
33
bench_parser = subparsers.add_parser(
34
"bench",
35
help="vLLM bench subcommand.",
36
+ description="vLLM bench subcommand.",
37
usage="vllm bench <bench_type> [options]")
38
bench_subparsers = bench_parser.add_subparsers(required=True,
39
dest="bench_type")
vllm/entrypoints/cli/openai.py
+5
-2
@@ -126,7 +126,8 @@ class ChatCommand(CLISubcommand):
126
subparsers: argparse._SubParsersAction) -> FlexibleArgumentParser:
127
chat_parser = subparsers.add_parser(
128
"chat",
129
- help="Generate chat completions via the running API server",
129
+ help="Generate chat completions via the running API server.",
130
+ description="Generate chat completions via the running API server.",
131
usage="vllm chat [options]")
132
_add_query_options(chat_parser)
133
chat_parser.add_argument(
@@ -162,7 +163,9 @@ class CompleteCommand(CLISubcommand):
163
complete_parser = subparsers.add_parser(
164
"complete",
165
help=("Generate text completions based on the given prompt "
165
- "via the running API server"),
166
+ "via the running API server."),
167
+ description=("Generate text completions based on the given prompt "
168
+ "via the running API server."),
169
usage="vllm complete [options]")
170
_add_query_options(complete_parser)
171
return complete_parser
vllm/entrypoints/cli/serve.py
+2
-1
@@ -34,7 +34,8 @@ class ServeSubcommand(CLISubcommand):
34
subparsers: argparse._SubParsersAction) -> FlexibleArgumentParser:
35
serve_parser = subparsers.add_parser(
36
"serve",
37
- help="Start the vLLM OpenAI Compatible API server",
37
+ help="Start the vLLM OpenAI Compatible API server.",
38
+ description="Start the vLLM OpenAI Compatible API server.",
39
usage="vllm serve [model_tag] [options]")
40
serve_parser.add_argument("model_tag",
41
type=str,