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

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2026-02-02 12:38:49 +00:00
committed by GitHub
parent b398e5c819
commit b10d05b8a8
8 changed files with 82 additions and 66 deletions
@@ -4,7 +4,7 @@
import asyncio
import math
from collections.abc import AsyncGenerator, Mapping
from typing import Literal, cast
from typing import Literal
import numpy as np
import torch
@@ -453,7 +453,10 @@ class VoxtralRealtimeGeneration(VoxtralForConditionalGeneration, SupportsRealtim
)
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)
},
)