herbiel

can we make multi-label ,such as 0,1,2,3....

i try on fs_cli console by /event CUSTOM google_transcribe::transcription, it can return result,but this.onceAsync('CUSTOM google_transcribe::transcription').then( function () { util.log('google_transcribe'); }); this.event_json('CUSTOM google_transcribe::transcription'); it can't work

yes ,i have get uuid ,but now i want to get event ,but still can't ,can you help me ? this is my event Event-Subclass: google_transcribe%3A%3Atranscription Event-Name: CUSTOM Core-UUID: 0a872b4e-f2b4-4096-90db-57782afa4e39...

now mod_google_transcribe can support check vad_detected event,if vad_detected has been checked ,it will trigger trancribe ,but if no voice ,if will keep wait for trancribe,i think may be it can...

> That log means that the environment variable has not been set -- at least not for the freeswitch process. > > The best place to set the environment variable...

yes, but i can't find any transcribe_audio function in the whisper or whisper-at

yes ,thanks you for you help ,but in the train stage,it still fail , File "/opt/whisper/whisper-at/src/whisper_at_train/models.py", line 172, in forward audio_rep = audio_rep.reshape(B*self.n_layer, audio_rep.shape[2], audio_rep.shape[3]) # [B*32, 25, 1280] RuntimeError:...

last answer can fix this problem. for option 2, please pip -e install our provided package. The transcribe_audio is at https://github.com/YuanGongND/whisper-at/blob/main/src/noise_robust_asr/intermediate_feat_extract/whisper_feat_extracrt/whisper/transcribe.py

just install from the https://github.com/YuanGongND/whisper-at/blob/main/src/noise_robust_asr/intermediate_feat_extract/whisper_feat_extracrt/whisper/ ,it will be normal