llama-bench: add -d
depth arg (#13096)
#13250
server.yml
on: push
server-windows
6m 15s
Matrix: server