Update app.py

app.py

@@ -8,19 +8,24 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(model_name)
 
 def generate_response(prompt):
-    input_ids = tokenizer.encode(prompt, return_tensors="pt")
+    instruction = f"Ответь кратко и ясно на вопрос:\n{prompt.strip()}\nОтвет:"
+    input_ids = tokenizer.encode(instruction, return_tensors="pt")
+
     output = model.generate(
         input_ids,
-        max_length=200,
-        num_return_sequences=1,
+        max_new_tokens=100,
         do_sample=True,
         top_k=50,
         top_p=0.95,
         temperature=0.9,
-        pad_token_id=tokenizer.eos_token_id
+        pad_token_id=tokenizer.eos_token_id,
+        eos_token_id=tokenizer.eos_token_id
     )
+
     response = tokenizer.decode(output[0], skip_special_tokens=True)
-    return response
+    
+    # Удалим часть промпта из начала
+    return response.replace(instruction, "").strip()
 
 iface = gr.Interface(
     fn=generate_response,