Update app.py

app.py

@@ -8,7 +8,7 @@ model = AutoModelForCausalLM.from_pretrained("BirdL/DeepSeek-Coder-V2-Lite-Instr
 
 @spaces.GPU
 def respond(message, history):
-    inputs = tokenizer(message, return_tensors="pt").to(model.device)
+    inputs = tokenizer(message, return_tensors="pt").input_ids.to("cuda")
     outputs = model.generate(inputs, max_new_tokens=192, do_sample=False, top_k=50, top_p=0.95, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
     ouputs = (tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))
     return outputs