#4308 TaskCfgSTT(is_cuda=True, uuid='793e70d1b5', cache_folder='E:/ytrans/tmp/13092/793e70d1b5', target_dir='E:/ytrans/output/

202.150* Posted at: 11 hours ago 👁14

ASR Error [faster-whisper (Local)] Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 257, in faster_whisper
File "faster_whisper\transcribe.py", line 586, in _batched_segments_generator
File "faster_whisper\transcribe.py", line 120, in forward
File "faster_whisper\transcribe.py", line 209, in generate_segment_batched
File "faster_whisper\transcribe.py", line 1400, in encode
RuntimeError: CUDA failed with error out of memory

Traceback (most recent call last):
File "videotrans\task\job.py", line 106, in run
File "videotrans\task\_speech2text.py", line 153, in recogn
File "videotrans\recognition\__init__.py", line 266, in run
File "videotrans\recognition\_base.py", line 142, in run
File "videotrans\recognition\_overall.py", line 63, in _exec
File "videotrans\recognition\_overall.py", line 136, in _faster
File "videotrans\configure\_base.py", line 290, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 257, in faster_whisper
File "faster_whisper\transcribe.py", line 586, in _batched_segments_generator
File "faster_whisper\transcribe.py", line 120, in forward
File "faster_whisper\transcribe.py", line 209, in generate_segment_batched
File "faster_whisper\transcribe.py", line 1400, in encode
RuntimeError: CUDA failed with error out of memory
TaskCfgSTT(is_cuda=True, uuid='793e70d1b5', cache_folder='E:/ytrans/tmp/13092/793e70d1b5', target_dir='E:/ytrans/output/recogn', source_language=None, source_language_code=None, source_sub=None, source_wav=None, source_wav_output=None, target_language=None, target_language_code=None, target_sub='E:/ytrans/output/recogn/1. What is Background Concept Art.srt', target_wav=None, target_wav_output=None, name='E:/2023/personal/inspo2/Redhong/Sect 01 Creating Concept Art for Game Backgrounds/1. What is Background Concept Art.mp4', noextname='1. What is Background Concept Art', basename='1. What is Background Concept Art.mp4', ext='mp4', dirname='E:/2023/personal/inspo2/Redhong/Sect 01 Creating Concept Art for Game Backgrounds', shound_del_name=None, detect_language='ko', recogn_type=0, model_name='large-v3-turbo', shibie_audio='E:/ytrans/tmp/13092/793e70d1b5/1. What is Background Concept Art-1777253079.109062.wav', remove_noise=False, enable_diariz=False, nums_diariz=0, rephrase=1, fix_punc=False)

=
system:Windows-10-10.0.26200-SP0
version:v3.97
frozen:True
language:en
root_dir:E:/ytrans

1 Replies

Floor #1 DeepSeek AI reply at 11 hours ago

CUDA out of memory during faster-whisper speech recognition, because the large-v3-turbo model requires more VRAM than your GPU can provide (minimum ~8–10GB for this model, and yours is insufficient).

Solutions (try in order):

  1. Disable "Secondary Recognition" if enabled on the main interface (top-right corner) – it doubles VRAM usage.
  2. Switch to a smaller model: choose medium, small, or base instead of large-v3-turbo. medium is a good balance.

  3. Lower VRAM consumption via advanced settings (go to Menu → Tools/Options → Advanced Options → Voice Recognition Parameters):

    • Set CUDA data type to int8 (or float16).
    • Set beam_size to 1.
    • Set best_of to 1.
    • Turn context off.
  4. Enable "Whisper pre‑split audio" in the same advanced panel – it processes smaller chunks, reducing peak VRAM.
  5. If VRAM is still too low (e.g., <4GB), disable CUDA entirely (uncheck it) and use CPU mode, though it will be very slow.

After making changes, click "Clean cached" and re‑run the task.

请查阅相关文档：

Post Your Reply
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.

Related resource