it has some problems on README

[Cyl147369991]

Tested versions

syntax.
FFmpeg version 5: Could not find module 'D:\soft\miniconda3\envs\speaker-split\Lib\site-packages\torchcodec\libtorchcodec_core5.dll' (or one of its dependencies). Try using the full path with constructor syntax.
FFmpeg version 4: Could not find module 'D:\soft\miniconda3\envs\speaker-split\Lib\site-packages\torchcodec\libtorchcodec_core4.dll' (or one of its dependencies). Try using the full path with constructor syntax.
[end of libtorchcodec loading traceback].
warnings.warn(
D:\soft\miniconda3\envs\speaker-split\lib\site-packages\speechbrain\utils\torch_audio_backend.py:57: UserWarning: torchaudio._backend.list_audio_backends has been deprecated. This deprecation is part of a large refactoring effort to transition TorchAudio into a maintenance phase. The decoding and encoding capabilities of PyTorch for both audio and video are being consolidated into TorchCodec. Please see pytorch/audio#3902 for more information. It will be removed from the 2.9 release.
available_backends = torchaudio.list_audio_backends()
Traceback (most recent call last):
File "e:\publicproject\CosyVoice\split-audio.py", line 14, in
diarization = pipeline("three.wav")
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\pipeline.py", line 476, in call
track_pipeline_apply(self, file, **kwargs)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\telemetry\metrics.py", line 152, in track_pipeline_apply
duration: float = Audio().get_duration(file)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\io.py", line 273, in get_duration
metadata: AudioStreamMetadata = get_audio_metadata(file)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\io.py", line 86, in get_audio_metadata
metadata = AudioDecoder(file["audio"]).metadata
NameError: name 'AudioDecoder' is not defined

System information

windows 11 pyannote.audio

Issue description

yntax.
FFmpeg version 5: Could not find module 'D:\soft\miniconda3\envs\speaker-split\Lib\site-packages\torchcodec\libtorchcodec_core5.dll' (or one of its dependencies). Try using the full path with constructor syntax.
FFmpeg version 4: Could not find module 'D:\soft\miniconda3\envs\speaker-split\Lib\site-packages\torchcodec\libtorchcodec_core4.dll' (or one of its dependencies). Try using the full path with constructor syntax.
[end of libtorchcodec loading traceback].
warnings.warn(
D:\soft\miniconda3\envs\speaker-split\lib\site-packages\speechbrain\utils\torch_audio_backend.py:57: UserWarning: torchaudio._backend.list_audio_backends has been deprecated. This deprecation is part of a large refactoring effort to transition TorchAudio into a maintenance phase. The decoding and encoding capabilities of PyTorch for both audio and video are being consolidated into TorchCodec. Please see pytorch/audio#3902 for more information. It will be removed from the 2.9 release.
available_backends = torchaudio.list_audio_backends()
Traceback (most recent call last):
File "e:\publicproject\CosyVoice\split-audio.py", line 14, in
diarization = pipeline("three.wav")
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\pipeline.py", line 476, in call
track_pipeline_apply(self, file, **kwargs)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\telemetry\metrics.py", line 152, in track_pipeline_apply
duration: float = Audio().get_duration(file)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\io.py", line 273, in get_duration
metadata: AudioStreamMetadata = get_audio_metadata(file)
File "D:\soft\miniconda3\envs\speaker-split\lib\site-packages\pyannote\audio\core\io.py", line 86, in get_audio_metadata
metadata = AudioDecoder(file["audio"]).metadata
NameError: name 'AudioDecoder' is not defined

Minimal reproduction example (MRE)

instantiate the pipeline from pyannote.audio import Pipeline import torch import os import sys sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(file)))) pipeline = Pipeline.from_pretrained( "pyannote/speaker-diarization-3.1", token="xxx" ) # pipeline.to(torch.device("cuda")) # run the pipeline on an audio file diarization = pipeline("three.wav") # dump the diarization output to disk using RTTM format with open("three.rttm", "w") as rttm: diarization.write_rttm(rttm)

[kkk112]

#I hope this can help, code was fixed by claude AI

import torch
import soundfile as sf
from pyannote.audio import Pipeline
from pyannote.audio.pipelines.utils.hook import ProgressHook

pipeline = Pipeline.from_pretrained(
"pyannote/speaker-diarization-3.1",
token="XXXXX")

pipeline.to(torch.device("cuda"))

audio_path = "Law.mp3"
waveform, sample_rate = sf.read(audio_path, always_2d=False, dtype='float32')

if waveform.ndim == 1:
waveform = waveform[None, :] # shape: (1, time)
else:
waveform = waveform.T # shape: (channels, time)

audio_input = {
"waveform": torch.from_numpy(waveform).float(),
"sample_rate": sample_rate
}

with ProgressHook() as hook:
output = pipeline(audio_input, hook=hook) # runs locally

for turn, speaker in output.speaker_diarization:
print(f"start={turn.start:.1f}s stop={turn.end:.1f}s speaker_{speaker}")

[Cyl147369991]

#I hope this can help, code was fixed by claude AI

import torch import soundfile as sf from pyannote.audio import Pipeline from pyannote.audio.pipelines.utils.hook import ProgressHook

pipeline = Pipeline.from_pretrained( "pyannote/speaker-diarization-3.1", token="XXXXX")

pipeline.to(torch.device("cuda"))

audio_path = "Law.mp3" waveform, sample_rate = sf.read(audio_path, always_2d=False, dtype='float32')

if waveform.ndim == 1: waveform = waveform[None, :] # shape: (1, time) else: waveform = waveform.T # shape: (channels, time)

audio_input = { "waveform": torch.from_numpy(waveform).float(), "sample_rate": sample_rate }

with ProgressHook() as hook: output = pipeline(audio_input, hook=hook) # runs locally

for turn, speaker in output.speaker_diarization: print(f"start={turn.start:.1f}s stop={turn.end:.1f}s speaker_{speaker}")

it works ,thank you very much