In request #31 I explained my issues trying to start a server, and over time of diagnosing I found that it actually had crashed, thus not allowing vector to connect to it. Restarting it causes it to crash over and over, but I think I've found the root cause of this crash, it's PicoVoice. Any time I have PicoVoice enabled, it constantly crashes, this is the error code:
Feb 08 02:25:56 escapepod start.sh[1392]: Starting webserver at port 8080 (http://localhost:8080)
Feb 08 02:25:56 escapepod start.sh[1392]: Starting SDK app
Feb 08 02:25:56 escapepod start.sh[1392]: Starting server at port 80 for connCheck
Feb 08 02:25:56 escapepod start.sh[1392]: Initiating leopard voice processor with language en-US
Feb 08 02:25:56 escapepod start.sh[1392]: Initializing 3 Picovoice Instances...
Feb 08 02:25:56 escapepod start.sh[1392]: Initializing Picovoice Instance 0
Feb 08 02:25:59 escapepod start.sh[1392]: Starting jdocs pinger timer for 192.168.1.6:443
Feb 08 02:25:59 escapepod start.sh[1392]: Using WofupOkqmmHZUhrUZvtMHA==
Feb 08 02:25:59 escapepod start.sh[1392]: Using WofupOkqmmHZUhrUZvtMHA==
Feb 08 02:26:00 escapepod start.sh[1392]: Error pinging jdocs, likely unauthenticated
Feb 08 02:26:06 escapepod start.sh[1392]: Initializing Picovoice Instance 1
Feb 08 02:26:16 escapepod start.sh[1392]: creating useepod
Feb 08 02:26:17 escapepod start.sh[1392]: Initiating vosk voice processor with language en-US
Feb 08 02:26:17 escapepod start.sh[1392]: Opening model
Feb 08 02:26:17 escapepod start.sh[1392]: Initializing Picovoice Instance 2
Feb 08 02:26:18 escapepod start.sh[1392]: LOG (VoskAPI:ReadDataFiles():model.cc:213) Decoding params beam=10 max-active=3000 lattice-beam=2
Feb 08 02:26:18 escapepod start.sh[1392]: LOG (VoskAPI:ReadDataFiles():model.cc:216) Silence phones 1:2:3:4:5:6:7:8:9:10
Feb 08 02:26:31 escapepod kernel: [ 1391] 0 1391 1674 0 40960 65 0 start.sh
Feb 08 02:26:31 escapepod start.sh[1391]: /home/pi/wire-prod-pod/chipper/start.sh: line 40: 1392 Killed ./chipper
For more context, if I use the recommended STT, it works perfectly fine, I know this because when I restart my server, it allows me to access the 8080 page for a short moment, allowing me to save the default setup settings, just enough time so the server does not crash, sometimes I have to do this more than once if I was not fast enough, after that I have tried a few different settings, majority of them work except for PicoVoice and also OpenAI, open AI seems to give a weird error when asking any question, so I just went back to Houndify, this is the error:
Feb 08 02:03:27 escapepod start.sh[4755]: Bot 1 Intent Sent: intent_knowledge_promptquestion
Feb 08 02:03:27 escapepod start.sh[4755]: No Parameters Sent
Feb 08 02:03:27 escapepod start.sh[4755]: Bot 1 request served.
Feb 08 02:03:30 escapepod start.sh[4755]: Bot 1 Stream type: OPUS
Feb 08 02:03:30 escapepod start.sh[4755]: (Bot 1, Vosk) Processing...
Feb 08 02:03:38 escapepod start.sh[4755]: (Bot 1) End of speech detected.
Feb 08 02:03:38 escapepod start.sh[4755]: Bot 1 Transcribed text: who is the vice president of america
Feb 08 02:03:39 escapepod start.sh[4755]: Making request to OpenAI...
Feb 08 02:03:40 escapepod start.sh[4755]: ERROR: {
Feb 08 02:03:40 escapepod start.sh[4755]: "error": {
Feb 08 02:03:40 escapepod start.sh[4755]: "message": "The server had an error while processing your request. Sorry about that!",
Feb 08 02:03:40 escapepod start.sh[4755]: "type": "server_error",
Feb 08 02:03:40 escapepod start.sh[4755]: "param": null,
Feb 08 02:03:40 escapepod start.sh[4755]: "code": null
Feb 08 02:03:40 escapepod start.sh[4755]: }
Feb 08 02:03:40 escapepod start.sh[4755]: }
In request #31 I explained my issues trying to start a server, and over time of diagnosing I found that it actually had crashed, thus not allowing vector to connect to it. Restarting it causes it to crash over and over, but I think I've found the root cause of this crash, it's PicoVoice. Any time I have PicoVoice enabled, it constantly crashes, this is the error code:
For more context, if I use the recommended STT, it works perfectly fine, I know this because when I restart my server, it allows me to access the 8080 page for a short moment, allowing me to save the default setup settings, just enough time so the server does not crash, sometimes I have to do this more than once if I was not fast enough, after that I have tried a few different settings, majority of them work except for PicoVoice and also OpenAI, open AI seems to give a weird error when asking any question, so I just went back to Houndify, this is the error: