Update app.py

app.py

@@ -9,7 +9,7 @@ import torch
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
 # Hugging Face Hub repository containing the fine-tuned model
-MODEL_REPO = "degide/tinyllama-medical-qa-assistant"
+MODEL_REPO = "degide/tinyllama-medical-assistant"
 
 print("Downloading and loading the fine-tuned medical chatbot...")
 
@@ -63,7 +63,7 @@ def generate_medical_response(message, history):
     
     prompt = (
         f"</s><|system|>You are a highly accurate and helpful medical assistant.</s>"
-        f"</s><|user|>{message}</s>"
+        f"<|user|>{message}</s>"
     )
     inputs = tokenizer(prompt, return_tensors="pt")
     
@@ -71,12 +71,10 @@ def generate_medical_response(message, history):
         outputs = model.generate(
             **inputs,
             max_new_tokens=256,
-            temperature=0.3, 
-            top_p=0.85,
+            temperature=0.3,
             repetition_penalty=1.0,
             do_sample=True,
-            eos_token_id=tokenizer.eos_token_id,
-            pad_token_id=tokenizer.eos_token_id
+            pad_token_id=tokenizer.eos_token_id,
         )
     
     generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)