[Model] Use explicit types in get_generation_prompt (#33551)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/model_executor/models/voxtral.py

@@ -25,7 +25,7 @@ from transformers.tokenization_utils_base import TextInput
 
 from vllm.config import ModelConfig, SpeechToTextConfig, VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
-from vllm.inputs.data import PromptType
+from vllm.inputs.data import PromptType, TokensPrompt
 from vllm.logger import init_logger
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.model_executor.model_loader.weight_utils import default_weight_loader
@@ -488,10 +488,13 @@ class VoxtralForConditionalGeneration(
         )
 
         tokenized = tokenizer.instruct.encode_transcription(req)
-        audio = (tokenized.audios[0].audio_array, stt_config.sample_rate)
-        prompts_dict = {"multi_modal_data": {"audio": audio}}
-        prompts_dict["prompt_token_ids"] = tokenized.tokens
-        return cast(PromptType, prompts_dict)
+
+        return TokensPrompt(
+            prompt_token_ids=tokenized.tokens,
+            multi_modal_data={
+                "audio": (tokenized.audios[0].audio_array, stt_config.sample_rate)
+            },
+        )
 
     @classmethod
     def get_num_audio_tokens(