Update run.py

run.py

@@ -24,8 +24,8 @@ from huggingface_hub import InferenceClient #multimodal_response
 # Specify models for text generation and embeddings
 #---------------------------------------------------
 
-#myModel="mistralai/Mixtral-8x7b-instruct-v0.1"
-myModel="princeton-nlp/gemma-2-9b-it-SimPO"
+myModel="mistralai/Mixtral-8x7b-instruct-v0.1"
+#myModel="princeton-nlp/gemma-2-9b-it-SimPO"
 #mod="mistralai/Mixtral-8x7b-instruct-v0.1"
 #tok=AutoTokenizer.from_pretrained(mod) #,token="hf_...")
 #cha=[{"role":"system","content":"A"},{"role":"user","content":"B"},{"role":"assistant","content":"C"}]
@@ -82,8 +82,8 @@ def format_prompt0(message, history):
 #-------------------------------------------------------------------------
 
 def format_prompt(message, history=None, system=None, RAGAddon=None, system2=None, zeichenlimit=None,historylimit=4, removeHTML=False,
-  startOfString="",template0="<start_of_turn>user\n{system}<end_of_turn>\n<start_of_turn>model\n<end_of_turn>\n",template1="<start_of_turn>user\n{message}<end_of_turn>\n<start_of_turn>model\n",template2="<end_of_turn>\n"):
-  #startOfString="<s>", template0=" [INST] {system} [/INST] </s>",template1=" [INST] {message} [/INST]",template2=" {response}</s>"):
+  #startOfString="",template0="<start_of_turn>user\n{system}<end_of_turn>\n<start_of_turn>model\n<end_of_turn>\n",template1="<start_of_turn>user\n{message}<end_of_turn>\n<start_of_turn>model\n",template2="<end_of_turn>\n"):
+  startOfString="<s>", template0=" [INST] {system} [/INST] </s>",template1=" [INST] {message} [/INST]",template2=" {response}</s>"):
   if zeichenlimit is None: zeichenlimit=1000000000 # :-)  
   prompt = ""
   if RAGAddon is not None: