#8880 Error during speech recognition stage[Audio.cpp(Local)] proc.returncode=1,stderr='load_backend: loaded CPU backend from

103.62**9 Posted at: 9 hours ago

Error during speech recognition stage[Audio.cpp(Local)] proc.returncode=1,stderr='load_backend: loaded CPU backend from C:\Users\user\Downloads\win-pyvideotrans-v4.19\audio_cpp\ggml-cpu-haswell.dll\naudiocpp_cli failed: Fun-ASR-Nano encoder input exceeds positional capacity\n'
stdout=''
cmd=['C:/Users/user/Downloads/win-pyvideotrans-v4.19/audio_cpp/audiocpp_cli', '--task', 'asr', '--family', 'fun_asr_nano', '--model', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/models/Fun-ASR-Nano-2512-GGUF/fun-asr-nano-2512-q8_0.gguf', '--backend', 'cpu', '--audio', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/clip_1795260767.2013018/audio_52.wav', '--language', 'en', '--request-option', 'audio_chunk_mode=none', '--text-out', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/audio_asr_txt_52.txt', '--threads', '8']
Traceback (most recent call last):
File "videotrans\task\only_one.py", line 85, in run
File "videotrans\task\_stage_recogn.py", line 64, in recogn
File "videotrans\recognition\__init__.py", line 178, in run
File "videotrans\recognition\_base.py", line 101, in run
File "videotrans\recognition\_audiocpp.py", line 80, in _exec
File "videotrans\util\help_misc.py", line 688, in exec_audiocpp
RuntimeError: proc.returncode=1,stderr='load_backend: loaded CPU backend from C:\Users\user\Downloads\win-pyvideotrans-v4.19\audio_cpp\ggml-cpu-haswell.dll\naudiocpp_cli failed: Fun-ASR-Nano encoder input exceeds positional capac
......
-v4.19/audio_cpp/audiocpp_cli', '--task', 'asr', '--family', 'fun_asr_nano', '--model', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/models/Fun-ASR-Nano-2512-GGUF/fun-asr-nano-2512-q8_0.gguf', '--backend', 'cpu', '--audio', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/clip_1795260767.2013018/audio_52.wav', '--language', 'en', '--request-option', 'audio_chunk_mode=none', '--text-out', 'C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/audio_asr_txt_52.txt', '--threads', '8']

trk=Translation V/A 单视频模式
Select Media File: C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p.mp4,
Output Directory: C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4,
self.cfg.cache_folder='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026'
Selected Clear Output, Not-Selected Enable accelerator
STT Service:Audio.cpp(Local), fun_asr_nano, Spoken: English
Max Speech Duration (s): 5s
Min Speech Duration (ms): 3000ms
Min Silence Duration for Split (ms): 600ms
VAD Speech Threshold: 0.45
VAD Engine: silero
TranslateSrt:Google Translate (Free),Spoken:English,Target:Bengali
Traditional Translation Batch Size (Lines): 5
TTS Service:Edge-TTS (Free), Voices:Tanishaa(Female/IN), Target:Bengali
Volume:+0%, Speed:+0%, Selected TTS Speedup, Selected Slow Video
Selected Trim Start/End Silence in Dubbed Segments
Silence Removal Intensity: default
Soft Subtitles
video duration:1709200ms
video information:{'video_fps': 25.0, 'r_frame_rate': 25.0, 'video_codec_name': 'h264', 'audio_codec_name': 'aac', 'width': 1920, 'height': 1080, 'time': 1709200, 'streams_len': 2, 'streams_audio': 1, 'video_streams': 1, 'color': 'yuv420p'}
self.cfg=TaskCfgVTT(uuid='2ca5f98026', name='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p.mp4', dirname='C:/Users/user/Downloads', noextname='She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p', basename='She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p.mp4', ext='mp4', target_dir='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4', cache_folder='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026', is_cuda=False, source_language='English', source_language_code='en', source_sub='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4/en.srt', source_wav='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/en.wav', source_wav_output='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4/en.m4a', target_language='Bengali', target_language_code='bn', target_sub='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4/bn.srt', target_wav='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/target.wav', target_wav_output='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4/bn.m4a', translate_type=0, tts_type=0, volume='+0%', pitch='+0Hz', voice_rate='+0%', voice_role='Tanishaa(Female/IN)', voice_autorate=True, video_autorate=True, remove_silent_mid=False, align_sub_audio=True, tts_model_name=None, detect_language='en', recogn_type=11, model_name='fun_asr_nano', shibie_audio=None, remove_noise=False, enable_diariz=False, nums_diariz=0, rephrase=False, fix_punc=0, app_mode='biaozhun', subtitle_language=None, targetdir_mp4='C:/Users/user/Downloads/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p-mp4/She Gave Birth to a Dragon The Mothers Monster Full Fantasy Movie 4K_1080p.mp4', novoice_mp4='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/novoice.mp4', is_separate=False, embed_bgm=True, instrument='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/instrument.wav', vocal='C:/Users/user/Downloads/win-pyvideotrans-v4.19/tmp/18416/2ca5f98026/vocal.wav', clear_cache=True, background_music='', subtitle_type=2, only_out_mp4=False, recogn2pass=False, output_srt=0, copysrt_rawvideo=False, loop_backaudio=1, backaudio_volume=0.8, batch=False)
video_duration_ms=1709200


=
system:Windows-10-10.0.26300-SP0
version:v4.19
frozen:True
language:en_US
root_dir:C:/Users/user/Downloads/win-pyvideotrans-v4.19

1 Replies

#1 DeepSeek reply at 9 hours ago

This is caused by the Audio.cpp channel's Fun-ASR-Nano model: the input audio clip exceeds the model's maximum position encoding length (positional capacity). The command line in the log contains --request-option audio_chunk_mode=none, meaning no chunking was applied, so a single VAD-segmented clip that is too long is sent directly to the model, and the model rejects it with an error.

Handling steps (in order of priority):

  1. Switch the STT channel (most reliable): change STT Service from Audio.cpp(Local) to faster-whisper(Built-in), and select the large-v3-turbo model (use large-v3 if the video has a lot of content). Keep Spoken as English.
  2. If you must use Audio.cpp: go to Menu → Tools/Options → Advanced Options → Speech Recognition (ASR), and reduce Max Speech Duration (s) from 5 to 2~3, so that each audio clip is shorter and does not exceed the model's capacity; then re-run and check Clear Output.
  3. Update the audio.cpp binary: delete the audiocpp_cli.exe (or audiocpp_cli) file in the software directory/audio_cpp/ folder, then run again. The latest version will be downloaded automatically; older versions have imperfect handling of long audio.
  4. Additionally, rename that video to a short English name with no spaces and move it out of the Downloads folder (for example D:/videotrans/001.mp4), then re-run. The current file name is extremely long and contains spaces, which can easily cause other errors in the subsequent FFmpeg stages.

If the error still occurs after switching to faster-whisper, provide about the last 30 lines of the error from the logs/ folder in the software installation directory for further analysis.

Please refer to the related documentation:

Thinking...
Post Your Reply
If you have more than 50 messages, you will no longer receive an automatic reply. Please ask a new question!
Open source and free maintenance is not easy. If this project is helpful to you, please consider making a small donation to help the project continue to maintain and update.

Dondate via ko-fi.com

Related resource