Update app.py

app.py

@@ -70,7 +70,7 @@ def transcribe(audio: tuple[int, np.ndarray], transformers_convo: list[dict], gr
         inputs["input_ids"] = inputs["input_ids"].to("cuda:0")
 
         generate_ids = model.generate(**inputs, max_length=256)
-        generate_ids = generate_ids[:, inputs.input_ids.size(1):]
+        generate_ids = generate_ids[:, inputs["input_ids"].size(1):]
         response = processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
         print("response", response)
         transformers_convo.append({"role": "assistant", "content": response})