Update app.py

app.py

@@ -3,7 +3,6 @@ import torch
 import gradio as gr
 import os
 
-hf_token = os.environ.get('HUGGINGFACE_TOKEN')
 
 # Define the device
 device = "cuda" if torch.cuda.is_available() else "cpu"
@@ -21,10 +20,9 @@ model.to(device)
 class ChatBot:
     def __init__(self):
         self.history = []
-    def predict(self, user_input, system_prompt="You are an expert medical analyst:"):
-        formatted_prompt = f"{user_input}\n{system_prompt}:"
-        prompt = [{'role': 'user', 'content': formatted_prompt}]
-        prompt = [{'role': 'user', 'content': user_input "\n" system_prompt}]
+
+    def predict(self, user_input, system_prompt="You are an expert analyst and provide assessment:"):
+        prompt = [{'role': 'user', 'content': user_input + "\n" + system_prompt + ":"}]
         inputs = tokenizer.apply_chat_template(
             prompt,
             add_generation_prompt=True,
@@ -39,13 +37,13 @@ class ChatBot:
             do_sample=False
         )
 
-        # Decode and return the response
+        # Decode the response
         response_text = tokenizer.decode(tokens[0], skip_special_tokens=False)
-        return response_text
-        
+
         # Free up memory
         del tokens
         torch.cuda.empty_cache()
+
         return response_text
 
 bot = ChatBot()
@@ -63,7 +61,7 @@ iface = gr.Interface(
     title=title,
     description=description,
     examples=examples,
-    inputs=["text", "text"],  # Take user input and system prompt separately
+    inputs=["text", "text"],
     outputs="text",
     theme="ParityError/Anime"
 )