我最近把一个用来生成SRT字幕文件的脚本从旧笔记本迁移到了新笔记本上，两边安装的依赖包版本完全一致，但在新笔记本的Anaconda独立环境里运行脚本时，却一直报找不到文件的错误，旧笔记本上运行完全正常，实在搞不懂哪里出问题了，希望大家能帮我看看。

错误回溯信息 #

Traceback (most recent call last):
  File "D:\Moives\OP\test.py", line 56, in <module>
    transcribe_audio(output_srt_path, file_name)
  File "D:\Moives\OP\test.py", line 11, in transcribe_audio
    transcribe_result = model.transcribe(audio=f"{file_name}", language="de")
  File "C:\Users\admin\anaconda3\lib\site-packages\whisper\transcribe.py", line 122, in transcribe
    mel = log_mel_spectrogram(audio, model.dims.n_mels, padding=N_SAMPLES)
  File "C:\Users\admin\anaconda3\lib\site-packages\whisper\audio.py", line 140, in log_mel_spectrogram
    audio = load_audio(audio)
  File "C:\Users\admin\anaconda3\lib\site-packages\whisper\audio.py", line 58, in load_audio
    out = run(cmd, capture_output=True, check=True).stdout
  File "C:\Users\admin\anaconda3\lib\subprocess.py", line 505, in run
    with Popen(*popenargs, **kwargs) as process:
  File "C:\Users\admin\anaconda3\lib\subprocess.py", line 951, in __init__
    self._execute_child(args, executable, preexec_fn, close_fds,
  File "C:\Users\admin\anaconda3\lib\subprocess.py", line 1420, in _execute_child
    hp, ht, pid, tid = _winapi.CreateProcess(executable, args,
FileNotFoundError: [WinError 2] The system cannot find the file specified

我的环境配置 #

• 操作系统：Windows 10
• Python版本：3.9.7
• openai-whisper版本：20231117
• argostranslate版本：1.9.6

我是在Anaconda创建的独立虚拟环境里运行的脚本。

已尝试的解决方法 #

• 给音频文件指定了完整路径
• 将文件名改成了无特殊字符的简单字符串
• 尝试升级了openai-whisper到更新的版本

但这些方法都没解决问题，旧笔记本上同样的脚本却能正常运行。

我的脚本代码 #

from datetime import timedelta
import os
import whisper
import argostranslate.package
import argostranslate.translate


def transcribe_audio(output_srt_path, file_name):
    model = whisper.load_model("small")  # Change this to your desired model
    
    transcribe_result = model.transcribe(audio=f"{file_name}", language="de")

    segments = transcribe_result['segments']

    germans = [segment['text'][1:] if segment['text'][0] == ' ' else segment['text'] for segment in segments ]
    translations = [argostranslate.translate.translate(sentence, "de", "en") for sentence in germans]

    with open(output_srt_path, 'w', encoding='utf-8') as srt_file:
        for idx, segment in enumerate(segments):
            try:
                start_time = str(timedelta(seconds=int(segment['start']))) + ',000'
                end_time = str(timedelta(seconds=int(segment['end']))) + ',000'

                text = segment['text']
                segment_id = segment['id'] + 1
                srt_segment = f"{segment_id}\n{start_time} --> {end_time}\n{text[1:] if text[0] == ' ' else text}\n"
                srt_file.write(srt_segment)
                
                text = ''.join(['-' for _ in range(min(80, len(segment['text']) + 10))])
                srt_segment = f"{text}\n"
                srt_file.write(srt_segment)
                
                english_translation = translations[segment['id']]
                srt_segment = f"{english_translation}\n\n"
                srt_file.write(srt_segment)
            except Exception as e:
                print(e)

    return output_srt_path

translated = set()
for file_name in os.listdir('./'):
    if file_name.endswith(".srt"):
        translated.add(f"{file_name[:-4]}")

for file_name in os.listdir('./'):
    if not file_name.endswith(".mp4"):
        continue
    if file_name in translated:
        continue

    # Example usage: Convert "audio.mp3" to an SRT file
    # don't forget to replace \n with a real new line
    output_srt_path = f"{file_name}.srt"

    transcribe_audio(output_srt_path, file_name)
    print(f"Transcription saved to {output_srt_path}")

备注：内容来源于stack exchange，提问作者laith drebaty