fix reference

scripts/run_server.py

@@ -1,4 +1,4 @@
-from server import ModelServer
+from model_server import ModelServer
 import argparse
 
 if __name__ == '__main__':

text-generation-webui/custom_model_server_monkey_patch.py

@@ -1,4 +1,4 @@
-from server import ModelClient
+from model_server import ModelClient
 from transformers import LlamaTokenizer
 
 def load_model_llama(*args, **kwargs):

text-generation-webui/generate_monkey_patch.py

@@ -1,6 +1,6 @@
 import modules.text_generation
 from modules.text_generation import *
-from server import _SentinelTokenStoppingCriteria
+from model_server import _SentinelTokenStoppingCriteria
 
 def generate_reply_patched(question, state, eos_token=None, stopping_strings=[]):
     if shared.model_name == 'None' or shared.model is None: