whisper.cpp icon indicating copy to clipboard operation
whisper.cpp copied to clipboard

server: --convert sometimes works

Open themanyone opened this issue 3 months ago • 1 comments

With the recent git update, the server --convert option no longer works properly all the time. After the second and subsequent access to the server, it responds with Enter command: |all

Tabbing over to the server. It has stopped and the ffmpeg implementation is asking whether or not to overwrite the temporary file, whisper_server_temp_file.wav_temp.wav. Saying yes to the prompt gets it to continue. But it happens frequently.

In the meantime, we are sending .wav files to the server, which have no problem since they do not rely on the --convert option or ffmpeg.

ffmpeg version 6.1.1 Copyright (c) 2000-2023 the FFmpeg developers built with gcc 13 (GCC) Input #0, mp3, from 'whisper_server_temp_file.wav': Metadata: encoder : Lavf60.16.100 Duration: 00:00:01.26, start: 0.069063, bitrate: 25 kb/s Stream #0:0: Audio: mp3, 16000 Hz, mono, fltp, 24 kb/s Stream mapping: Stream #0:0 -> #0:0 (mp3 (mp3float) -> pcm_s16le (native)) Press [q] to stop, [?] for help Output #0, wav, to 'whisper_server_temp_file.wav_temp.wav': Metadata: ISFT : Lavf60.16.100 Stream #0:0: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 16000 Hz, mono, s16, 256 kb/s Metadata: encoder : Lavc60.31.102 pcm_s16le [out#0/wav @ 0x55d9af2c6f00] video:0kB audio:37kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.207094% size= 37kB time=00:00:01.15 bitrate= 261.4kbits/s speed= 375x
Successfully loaded /tmp/tmpn3nh19qc.mp3 Running whisper.cpp inference on /tmp/tmpn3nh19qc.mp3

themanyone avatar Mar 23 '24 19:03 themanyone

Can't reproduce on MacOS

ggerganov avatar Mar 28 '24 09:03 ggerganov