41 lines
1.5 KiB
Python
41 lines
1.5 KiB
Python
|
import sys
|
|||
|
|
import logging
|
|||
|
|
|
|||
|
|
# 配置日志
|
|||
|
|
logging.basicConfig(level=logging.INFO)
|
|||
|
|
logger = logging.getLogger("whisper_patch")
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
import whisper as w
|
|||
|
|
logger.info("成功导入原始whisper库")
|
|||
|
|
original_whisper_available = True
|
|||
|
|
except ImportError:
|
|||
|
|
logger.warning("无法导入原始whisper库，将使用模拟实现")
|
|||
|
|
original_whisper_available = False
|
|||
|
|
# 创建一个空的模块作为替代
|
|||
|
|
class DummyWhisperModule:
|
|||
|
|
pass
|
|||
|
|
w = DummyWhisperModule()
|
|||
|
|
|
|||
|
|
class Model:
|
|||
|
|
@staticmethod
|
|||
|
|
def load_model(name, device="cpu", download_root=None):
|
|||
|
|
logger.info(f"使用whisper兼容层加载模型: {name}, 设备: {device}")
|
|||
|
|
return DummyModel()
|
|||
|
|
|
|||
|
|
class DummyModel:
|
|||
|
|
def transcribe(self, audio_path, **kwargs):
|
|||
|
|
logger.info(f"使用模拟转录功能处理音频: {audio_path}")
|
|||
|
|
segments = [
|
|||
|
|
{"text": "这是一个示例PPT讲解视频。", "start": 0.0, "end": 5.0},
|
|||
|
|
{"text": "由于Python版本限制，无法使用原始语音识别功能。", "start": 5.0, "end": 10.0},
|
|||
|
|
{"text": "这是一个兼容性替代方案。", "start": 10.0, "end": 15.0},
|
|||
|
|
{"text": "您可以升级到Python 3.8+以使用完整功能。", "start": 15.0, "end": 20.0}
|
|||
|
|
]
|
|||
|
|
return {"segments": segments}
|
|||
|
|
|
|||
|
|
# 将自定义功能添加到whisper模块
|
|||
|
|
sys.modules["whisper"] = w
|
|||
|
|
sys.modules["whisper"].load_model = Model.load_model
|
|||
|
|
|
|||
|
|
logger.info("whisper兼容层初始化完成")
|