try and avoid weird server test failure (spillage / parallelism between completion & tool call tests?)

Author: Olivier Chafik

examples/server/tests/unit/test_tool_call.py

@@ -10,6 +10,8 @@
 def create_server():
     global server
     server = ServerPreset.tinyllama2()
+    server.model_alias = "tinyllama-2-tool-call"
+    server.server_port = 8081
 
 
 TEST_TOOL = {