描述 #

AudioLidWhisper 音频语言识别处理器

核心功能 #

• 音频解码：支持多种格式音频 → 16kHz 波形数据
• 语言检测：基于Whisper-large模型识别几十种语言代码
• 全称映射：自动转换语言代码为完整语言名称

技术特性 #

• 基于ModelScope框架加载模型
• 16kHz采样率音频预处理流水线

典型应用场景 #

• ✅ 语音内容分析 - 多语种音频分类
• ✅ 流媒体处理 - 实时语言识别
• ✅ 语音数据集 - 自动化语言标注

建议 #

• 先将上游音频统一标准化成 wav 格式，再进行语种识别，有助于提升准确性。

算子参数 #

输入 #

输出 #

结构化数组，每个元素包含：

• language_code: 语言代码
• language_code_full_name: 语言完整英文名称

参数 #

如参数没有默认值，则为必填参数

调用示例 #

下面的代码展示了如何使用 daft 运行算子识别音频语种。

from __future__ import annotations

import logging
import os

import ray

import daft
from daft import col
from daft.las.functions.audio.audio_lid_whisper import AudioLidWhisper
from daft.las.functions.udf import las_udf

if __name__ == "__main__":
    if os.getenv("DAFT_RUNNER", "ray") == "ray":

        def configure_logging():
            logging.basicConfig(
                level=logging.INFO,
                format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
                datefmt="%Y-%m-%d %H:%M:%S.%s".format(),
            )
            logging.getLogger("tracing.span").setLevel(logging.WARNING)
            logging.getLogger("daft_io.stats").setLevel(logging.WARNING)
            logging.getLogger("DaftStatisticsManager").setLevel(logging.WARNING)
            logging.getLogger("DaftFlotillaScheduler").setLevel(logging.WARNING)
            logging.getLogger("DaftFlotillaDispatcher").setLevel(logging.WARNING)

        import ray

        ray.init(dashboard_host="0.0.0.0", runtime_env={"worker_process_setup_hook": configure_logging})
        daft.context.set_runner_ray()

    daft.set_execution_config(actor_udf_ready_timeout=600)
    daft.set_execution_config(min_cpu_per_task=0)

    tos_dir_url = os.getenv("TOS_DIR_URL", "las-cn-beijing-public-online.tos-cn-beijing.volces.com")
    samples = {
        "audio_path": [
            f"https://{tos_dir_url}/public/shared_audio_dataset/参观八达岭长城。.wav"
        ]
    }

    model_path = os.getenv("MODEL_PATH", "/opt/las/models")
    model_name = "iic/speech_whisper-large_lid_multilingual_pytorch"
    model_version = "v2.0.4"
    num_gpus = 1
    rank = 0

    df = daft.from_pydict(samples)
    df = df.with_column(
        "lid_result",
        las_udf(
            AudioLidWhisper,
            construct_args={
                "model_path": model_path,
                "model_name": model_name,
                "model_version": model_version,
                "rank": rank,
            },
            num_gpus=1,
            batch_size=1,
            concurrency=1,
        )(col("audio_path")),
    )
    df.show()

    # ╭────────────────────────────────┬────────────────────────────────────────────────────────────╮
    # │ audio_path                     ┆ lid_result                                                 │
    # │ ---                            ┆ ---                                                        │
    # │ Utf8                           ┆ Struct[language_code: Utf8, language_code_full_name: Utf8] │
    # ╞════════════════════════════════╪════════════════════════════════════════════════════════════╡
    # │ tos://las-cn-beijing-public-o… ┆ {language_code: zh,                                        │
    # │                                ┆ language_…                                                 │
    # ╰────────────────────────────────┴────────────────────────────────────────────────────────────╯