synesthesiam
commented
4 years ago rhasspy-server-hermes doesn't usually subscribe to audio frames. You need to set the "Satellite site IDs" in the settings page for your ASR (speech to text) and NLU (text to intent) systems.
After upgrading to 2.5 (coming from 2.5-pre) with the docker image rhasspyserver_hermes seems not to subscribe the siteIDs for recording audio frames. Config:
{ "dialogue": { "satellite_site_ids": "s1,s2,s3,s4,s5,s6", "system": "rhasspy" }, "intent": { "fuzzywuzzy": { "min_confidence": "0.2" }, "satellite_site_ids": "s1,s2,s3,s4,s5,s6", "system": "fuzzywuzzy" }, "microphone": { "system": "hermes" }, "mqtt": { "enabled": "true", "host": "rhasspy" }, "speech_to_text": { "satellite_site_ids": "s1,s2,s3,s4,s5,s6", "system": "pocketsphinx" }, "wake": { "pocketsphinx": { "keyphrase": "alexa" }, "porcupine": { "keyword_path": "alexa.ppn", "sensitivity": "0.5", "udp_audio": "" }, "satellite_site_ids": "s1,s2,s3,s4,s5,s6", "snowboy": { "model": "alexa.ppn" }, "system": "porcupine" } }Log:
[DEBUG:2020-06-09 19:08:32,380] rhasspyprofile.download: speech_to_text.deepspeech.mix_weight >0 0 = False [DEBUG:2020-06-09 19:08:32,380] rhasspyprofile.download: speech_to_text.kaldi.mix_weight >0 0 = False [DEBUG:2020-06-09 19:08:32,379] rhasspyprofile.download: speech_to_text.pocketsphinx.mix_weight >0 0 = False [DEBUG:2020-06-09 19:08:32,379] rhasspyprofile.download: speech_to_text.deepspeech.open_transcription True False = False [DEBUG:2020-06-09 19:08:32,379] rhasspyprofile.download: speech_to_text.kaldi.open_transcription True False = False [DEBUG:2020-06-09 19:08:32,378] rhasspyprofile.download: speech_to_text.pocketsphinx.open_transcription True False = False [DEBUG:2020-06-09 19:08:32,378] rhasspyprofile.download: speech_to_text.system deepspeech pocketsphinx = False [DEBUG:2020-06-09 19:08:32,378] rhasspyprofile.download: speech_to_text.system kaldi pocketsphinx = False [DEBUG:2020-06-09 19:08:32,378] rhasspyprofile.download: Skipping acoustic_model/variances (/profiles/de/acoustic_model/variances) [DEBUG:2020-06-09 19:08:32,378] rhasspyprofile.download: Skipping acoustic_model/transition_matrices (/profiles/de/acoustic_model/transition_matrices) [DEBUG:2020-06-09 19:08:32,377] rhasspyprofile.download: Skipping acoustic_model/noisedict (/profiles/de/acoustic_model/noisedict) [DEBUG:2020-06-09 19:08:32,377] rhasspyprofile.download: Skipping acoustic_model/mixture_weights (/profiles/de/acoustic_model/mixture_weights) [DEBUG:2020-06-09 19:08:32,377] rhasspyprofile.download: Skipping acoustic_model/means (/profiles/de/acoustic_model/means) [DEBUG:2020-06-09 19:08:32,376] rhasspyprofile.download: Skipping acoustic_model/mdef (/profiles/de/acoustic_model/mdef) [DEBUG:2020-06-09 19:08:32,376] rhasspyprofile.download: Skipping acoustic_model/feature_transform (/profiles/de/acoustic_model/feature_transform) [DEBUG:2020-06-09 19:08:32,376] rhasspyprofile.download: Skipping acoustic_model/feat.params (/profiles/de/acoustic_model/feat.params) [DEBUG:2020-06-09 19:08:32,376] rhasspyprofile.download: Skipping g2p.fst (/profiles/de/g2p.fst) [DEBUG:2020-06-09 19:08:32,375] rhasspyprofile.download: Skipping base_dictionary.txt (/profiles/de/base_dictionary.txt) [DEBUG:2020-06-09 19:08:32,375] rhasspyprofile.download: speech_to_text.system pocketsphinx pocketsphinx = True [INFO:2020-06-09 19:08:32,302] rhasspyserver_hermes: Started [DEBUG:2020-06-09 19:08:32,302] rhasspyserver_hermes: Subscribed to hermes/hotword/+/detected [DEBUG:2020-06-09 19:08:32,301] rhasspyserver_hermes: Subscribed to hermes/nlu/intentNotRecognized [DEBUG:2020-06-09 19:08:32,301] rhasspyserver_hermes: Subscribed to hermes/asr/textCaptured [DEBUG:2020-06-09 19:08:32,301] rhasspyserver_hermes: Subscribed to rhasspy/asr/default/default/audioCaptured [DEBUG:2020-06-09 19:08:32,300] rhasspyserver_hermes: Subscribed to hermes/intent/# [DEBUG:2020-06-09 19:08:32,300] rhasspyserver_hermes: Subscribed to hermes/audioServer/default/audioSummary [DEBUG:2020-06-09 19:08:32,300] rhasspyserver_hermes: Subscribed to hermes/audioServer/default/audioSummary [DEBUG:2020-06-09 19:08:32,300] rhasspyserver_hermes: Subscribed to rhasspy/asr/default/default/audioCaptured [DEBUG:2020-06-09 19:08:32,299] rhasspyserver_hermes: Subscribed to hermes/nlu/intentNotRecognized [DEBUG:2020-06-09 19:08:32,299] rhasspyserver_hermes: Subscribed to hermes/intent/# [DEBUG:2020-06-09 19:08:32,299] rhasspyserver_hermes: Subscribed to hermes/asr/textCaptured [DEBUG:2020-06-09 19:08:32,298] rhasspyserver_hermes: Subscribed to hermes/hotword/+/detected [DEBUG:2020-06-09 19:08:32,298] rhasspyserver_hermes: Connected to MQTT broker [DEBUG:2020-06-09 19:08:32,294] rhasspyserver_hermes: Connecting to rhasspy:1883 (retries: 0/10) [DEBUG:2020-06-09 19:08:32,294] rhasspyserver_hermes: Starting core [DEBUG:2020-06-09 19:08:32,291] rhasspyprofile.profile: Loading default profile settings from /usr/lib/rhasspy/lib/python3.7/site-packages/rhasspyprofile/profiles/defaults.json [DEBUG:2020-06-09 19:08:32,291] rhasspyprofile.profile: Loading /profiles/de/profile.json [DEBUG:2020-06-09 19:08:32,290] rhasspyprofile.profile: Loading /usr/lib/rhasspy/lib/python3.7/site-packages/rhasspyprofile/profiles/de/profile.json [DEBUG:2020-06-09 19:08:32,111] rhasspyserver_hermes: Restarting RhasspyI want to use multiple "stupid" pi zeros with only audio-streaming to my central VM and there is no wake-word detection even if i push the "Wake" button on the Web-GUI on my VM...