Running WhisperX on Jetson Nano Orin 8GB

Autonomous Machines Jetson & Embedded Systems Jetson Orin Nano
1

Hi, I just installed WhisperX successfully. However, when calling the example Python script I am getting the following error during whisperx package import:

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
Cell In[3], line 1
----> 1 import whisperx
      2 import gc 

File ~/Git/whisperX/whisperx/__init__.py:1
----> 1 from .transcribe import load_model
      2 from .alignment import load_align_model, align
      3 from .audio import load_audio

File ~/Git/whisperX/whisperx/transcribe.py:10
      7 import torch
      9 from .alignment import align, load_align_model
---> 10 from .asr import load_model
     11 from .audio import load_audio
     12 from .diarize import DiarizationPipeline, assign_word_speakers

File ~/Git/whisperX/whisperx/asr.py:5
      2 import warnings
      3 from typing import List, Union, Optional, NamedTuple
----> 5 import ctranslate2
      6 import faster_whisper
      7 import numpy as np

File ~/Git/whisperX/.venv/lib/python3.10/site-packages/ctranslate2/__init__.py:21
     18         ctypes.CDLL(library)
     20 try:
---> 21     from ctranslate2._ext import (
     22         AsyncGenerationResult,
     23         AsyncScoringResult,
     24         AsyncTranslationResult,
     25         DataType,
     26         Device,
     27         Encoder,
     28         EncoderForwardOutput,
     29         ExecutionStats,
     30         GenerationResult,
     31         GenerationStepResult,
     32         Generator,
     33         MpiInfo,
     34         ScoringResult,
     35         StorageView,
     36         TranslationResult,
     37         Translator,
     38         contains_model,
     39         get_cuda_device_count,
     40         get_supported_compute_types,
     41         set_random_seed,
     42     )
     43     from ctranslate2.extensions import register_extensions
     44     from ctranslate2.logging import get_log_level, set_log_level

ImportError: /home/username/Git/whisperX/.venv/lib/python3.10/site-packages/ctranslate2/../ctranslate2.libs/libgomp-d22c30c5.so.1.0.0: cannot allocate memory in static TLS block

I already tried “export LD_PRELOAD=/usr/lib/aarch64-linux-gnu/libgomp.so.1” but it didn’t help.

Any idea how to fix it?

3

Hi,

We can import whisperx normally although some user warnings.
Please find our steps below:

$ pip3 install whisperx
$ python3
Python 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import whisperx
/usr/lib/python3/dist-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.17.3 and <1.25.0 is required for this version of SciPy (detected version 1.26.4
  warnings.warn(f"A NumPy version >={np_minversion} and <{np_maxversion}"
/home/nvidia/.local/lib/python3.10/site-packages/pyannote/audio/core/io.py:43: UserWarning: torchaudio._backend.set_audio_backend has been deprecated. With dispatcher enabled, this function is no-op. You can remove the function call.
  torchaudio.set_audio_backend("soundfile")
/home/nvidia/.local/lib/python3.10/site-packages/pyannote/audio/pipelines/speaker_verification.py:43: UserWarning: torchaudio._backend.get_audio_backend has been deprecated. With dispatcher enabled, this function is no-op. You can remove the function call.
  backend = torchaudio.get_audio_backend()
/home/nvidia/.local/lib/python3.10/site-packages/pyannote/audio/pipelines/speaker_verification.py:45: UserWarning: Module 'speechbrain.pretrained' was deprecated, redirecting to 'speechbrain.inference'. Please update your script. This is a change from SpeechBrain 1.0. See: https://github.com/speechbrain/speechbrain/releases/tag/v1.0.0
  from speechbrain.pretrained import (
/home/nvidia/.local/lib/python3.10/site-packages/pyannote/audio/pipelines/speaker_verification.py:53: UserWarning: torchaudio._backend.set_audio_backend has been deprecated. With dispatcher enabled, this function is no-op. You can remove the function call.
  torchaudio.set_audio_backend(backend)
/home/nvidia/.local/lib/python3.10/site-packages/pyannote/audio/tasks/segmentation/mixins.py:37: UserWarning: `torchaudio.backend.common.AudioMetaData` has been moved to `torchaudio.AudioMetaData`. Please update the import path.
  from torchaudio.backend.common import AudioMetaData
>>>

Thanks.

4

Hi,

Thanks a lot for testing it out. Ok, I will reinstall JetPack because I wanna upgrade anyway. Maybe it will fix the issue.

5

Hi,

If you are still facing the same issue after reinstalling, please attach the WhisperX installation command wit us.

Thanks.