20 lines
554 B
Python
20 lines
554 B
Python
"""Launch the inference server."""
|
|
|
|
import os
|
|
import sys
|
|
|
|
from sglang.srt.server import launch_server
|
|
from sglang.srt.server_args import prepare_server_args
|
|
from sglang.srt.utils import kill_child_process
|
|
|
|
if __name__ == "__main__":
|
|
server_args = prepare_server_args(sys.argv[1:])
|
|
model_override_args = server_args.json_model_override_args
|
|
|
|
try:
|
|
launch_server(server_args, model_override_args=model_override_args)
|
|
except Exception as e:
|
|
raise e
|
|
finally:
|
|
kill_child_process(os.getpid(), including_parent=False)
|