Use the output of --live_transcribe

[fatualux]

Hi, and thanks for your project, it works flawlessly and I had met no issue at all. I would just like to ask if there is a way to "trap" or echo somewhere the real time transcriptions obtained with the option --live_transcribe True Thanks in advance for your support

[ArnaudKunzi]

Sounds like a job for tee.

[fatualux]

Thanks I did not know this package. I will try to write a sctipt to integrate tee with --live_transcribe.

[mpuz]

Hi, and thanks for your project, it works flawlessly and I had met no issue at all. I would just like to ask if there is a way to "trap" or echo somewhere the real time transcriptions obtained with the option --live_transcribe True Thanks in advance for your support

Looks like You are the happy one who succeed to have live-transcribe to work, can You share your cli args and what language do You transcribe?

for saving output You can feed it to the output to file - read help

whisper-ctranslate2 --help

take a look at

Configuration options to control generated outputs:
  --output_dir OUTPUT_DIR, -o OUTPUT_DIR
                        directory to save the outputs (default: .)
  --output_format {txt,vtt,srt,tsv,json,all}, -f {txt,vtt,srt,tsv,json,all}
                        format of the output file; if not specified, all
                        available formats will be produced (default: all)
  --pretty_json PRETTY_JSON, -p PRETTY_JSON
                        produce json in a human redable format (default:
                        False)

[fatualux]

yes, of course. I transcribe Italian... Here is the link to the script I am now using for live transcription. It is a bit slow in transcribing, so I always use small model.

https://github.com/fatualux/scripts/blob/main/realtimeWhisper.sh

..:FZ:..

• Sent from my Android device -

Il gio 19 ott 2023, 20:44 Michael @.***> ha scritto:

Hi, and thanks for your project, it works flawlessly and I had met no issue at all. I would just like to ask if there is a way to "trap" or echo somewhere the real time transcriptions obtained with the option --live_transcribe True Thanks in advance for your support

Looks like You are the happy one who succeed to have live-transcribe to work, can You share your cli args and what language do You transcribe?

— Reply to this email directly, view it on GitHub https://github.com/Softcatala/whisper-ctranslate2/issues/52#issuecomment-1771528271, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIPQJXGQCASMRWLG6MIUCVDYAFYI3AVCNFSM6AAAAAA2NRAHGGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONZRGUZDQMRXGE . You are receiving this because you authored the thread.Message ID: @.***>

[P3NG00]

Hi, and thanks for your project, it works flawlessly and I had met no issue at all. I would just like to ask if there is a way to "trap" or echo somewhere the real time transcriptions obtained with the option --live_transcribe True Thanks in advance for your support

Looks like You are the happy one who succeed to have live-transcribe to work, can You share your cli args and what language do You transcribe?

for saving output You can feed it to the output to file - read help

whisper-ctranslate2 --help

take a look at

Configuration options to control generated outputs:
  --output_dir OUTPUT_DIR, -o OUTPUT_DIR
                        directory to save the outputs (default: .)
  --output_format {txt,vtt,srt,tsv,json,all}, -f {txt,vtt,srt,tsv,json,all}
                        format of the output file; if not specified, all
                        available formats will be produced (default: all)
  --pretty_json PRETTY_JSON, -p PRETTY_JSON
                        produce json in a human redable format (default:
                        False)

I cannot get the output format or output dir to work for --live_transcribe. Tee barely works, I would like some way to just get the text output so that I can put it into another program.