add .to(0) for inputs

app/api.py

@@ -108,7 +108,7 @@ async def text_generate(
         max_penalty = 1.5
         repetition_penalty = max(min_penalty + (1.0 - temperature) * (max_penalty - min_penalty), 0.8)
     prompt = f"User: {text}\nAssistant: "
-    input_ids = text_generator[model_name]["tokenizer"](prompt, return_tensors='pt').input_ids  # .to(device)
+    input_ids = text_generator[model_name]["tokenizer"](prompt, return_tensors='pt').input_ids.to(0)
     text_generator[model_name]["model"].eval()
     print("Generating text...")
     print(f"max_length: {max_length}, do_sample: {do_sample}, top_k: {top_k}, top_p: {top_p}, "