Update app.py

app.py

@@ -51,7 +51,7 @@ def format_prompt(message, history):
     prompt += f"<start_of_turn>user{message}<end_of_turn><start_of_turn>model"
     print(prompt)
     return prompt
-mega_hist=[]
+mega_hist=[[],[],[],[]]
 
 
 def chat_inf(system_prompt,prompt,history,client_choice,seed,temp,tokens,top_p,rep_p,hid_val):
@@ -60,19 +60,19 @@ def chat_inf(system_prompt,prompt,history,client_choice,seed,temp,tokens,top_p,r
         #token max=8192
         client=client_z[int(hid_val)-1]
         
-        if not mega_hist[hid_val-1]:
-            mega_hist.append([])
+        if history:
+            mega_hist[hid_val-1]=history
             #history = []
             hist_len=0
-        if mega_hist[hid_val-1]:
-            hist_len=len(mega_hist[hid_val-1])
-            print(hist_len)
-        in_len=len(system_prompt+prompt)+hist_len
-        print("\n#########"+str(in_len))
-        if (in_len+tokens) > 8000:
-            yield [(prompt,"Wait. I need to compress our Chat history...")]
-            #history=compress_history(history,client_choice,seed,temp,tokens,top_p,rep_p)
-            yield [(prompt,"History has been compressed, processing request...")]
+        #if mega_hist[hid_val-1]:
+        #    hist_len=len(mega_hist[hid_val-1])
+        #    print(hist_len)
+        #in_len=len(system_prompt+prompt)+hist_len
+        #print("\n#########"+str(in_len))
+        #if (in_len+tokens) > 8000:
+        #    yield [(prompt,"Wait. I need to compress our Chat history...")]
+        #    #history=compress_history(history,client_choice,seed,temp,tokens,top_p,rep_p)
+        #    yield [(prompt,"History has been compressed, processing request...")]
                 
         generate_kwargs = dict(
             temperature=temp,