add tokenizer name config to set the vllm/sglang tokenizer to something different if needed

atroposlib/envs/server_handling/server_manager.py

@@ -121,6 +121,7 @@ class ServerManager:
                         model_name=configs.model_name,
                         rolling_buffer_length=configs.rolling_buffer_length,
                         api_key="x",
+                        tokenizer_name=configs.tokenizer_name,
                     )
                 )
             self.servers = [