Update app.py

app.py

@@ -36,7 +36,7 @@ model_name = "petals-team/StableBeluga2"
 #daekeun-ml/Llama-2-ko-DPO-13B
 #daekeun-ml/Llama-2-ko-instruct-13B
 #quantumaikr/llama-2-70b-fb16-korean
-tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
+#tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
 
 model = None
 
@@ -155,9 +155,10 @@ def init():
 
 
 def chat(id, npc, text):
-    if model == None:
-        init()
-        return "no model"
+    #if model == None:
+    #    init()
+    #    return "no model"
+    
     # get_coin endpoint
     response = requests.post("https://ldhldh-api-for-unity.hf.space/run/predict_6", json={
       "data": [
@@ -170,7 +171,7 @@ def chat(id, npc, text):
     
     # model inference
 
-    if check(model_name):
+    if True:
         
         global history
         if not npc in npc_story:
@@ -213,11 +214,12 @@ def chat(id, npc, text):
 ### {npc}:
 """
 
-        inputs = tokenizer(prom, return_tensors="pt")["input_ids"]
-        outputs = model.generate(inputs, do_sample=True, temperature=0.6, top_p=0.75, max_new_tokens=100)
-        output = tokenizer.decode(outputs[0])[len(prom)+3:-1].split("<")[0].split("###")[0].replace(". ", ".\n")
-        output = cleanText(output)
-        print(tokenizer.decode(outputs[0]))
+        #inputs = tokenizer(prom, return_tensors="pt")["input_ids"]
+        #outputs = model.generate(inputs, do_sample=True, temperature=0.6, top_p=0.75, max_new_tokens=100)
+        #output = tokenizer.decode(outputs[0])[len(prom)+3:-1].split("<")[0].split("###")[0].replace(". ", ".\n")
+        #output = cleanText(output)
+        #print(tokenizer.decode(outputs[0]))
+        output = f"{npc}의 응답입니다."
         print(output)
         history[npc][id] += f"\n\n### User:\n{text}\n\n### {npc}:{output}"
     else: