Update app.py

app.py

@@ -33,7 +33,7 @@ def generate(
 ):
     """run model inference, will return a Generator if streaming is true"""
 
-    return next(llm(format_prompt(user_prompt), **asdict(generation_config)))["text"]
+    return llm(format_prompt(user_prompt), **asdict(generation_config))
 
 config = AutoConfig.from_pretrained(
     "teknium/Replit-v2-CodeInstruct-3B", context_length=2048
@@ -70,8 +70,11 @@ examples = [example_1, example_2]
 
 def generate_code(user_input):
     response = generate(llm, generation_config, user_input)
-    print(response)
-    return response
+    code = ""
+    for word in response:
+        code = code + word
+    print(code)
+    return code
 
 UI = gr.Interface(
     fn=generate_code,