search

OpenVoiceOS / ovos-plugin-manager

plugin manager for OpenVoiceOS , STT/TTS/Wakewords that can be used anywhere
Apache License 2.0
12 stars 11 forks source link

An error occurred (InvalidSsmlException) when calling the SynthesizeSpeech operation: Invalid SSML request #221

Open goldyfruit opened 5 months ago

goldyfruit commented 5 months ago

While using fallback chatGPT skill with LLAMA 3, the skill returns Speak: <|im_start|>user which generates an error with the TTS.

Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: 2024-04-24 17:40:58.319 - HiveMind-voice-sat - ovos_audio.service:execute_tts:344 - INFO - Speak: <|im_start|>user
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: 2024-04-24 17:40:58.373 - HiveMind-voice-sat - ovos_audio.service:execute_tts:350 - ERROR - TTS synth failed! An error occurred (InvalidSsmlException) when calling the SynthesizeSpeech operation: Invalid SSML request
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: Traceback (most recent call last):
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/ovos_audio/service.py", line 347, in execute_tts
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     self.tts.execute(utterance, ident, listen,
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/ovos_plugin_manager/templates/tts.py", line 498, in execute
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     self._execute(sentence, ident, listen, **kwargs)
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/ovos_plugin_manager/templates/tts.py", line 527, in _execute
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     audio_file, phonemes = self.synth(sentence, **kwargs)
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/ovos_plugin_manager/templates/tts.py", line 587, in synth
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     audio.path, phonemes = self.get_tts(sentence, str(audio), **kwargs)
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/ovos_tts_plugin_polly/__init__.py", line 73, in get_tts
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     response = self.polly.synthesize_speech(
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/botocore/client.py", line 565, in _api_call
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     return self._make_api_call(operation_name, kwargs)
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:   File "/home/goldyfruit/.venvs/ovos/lib/python3.11/site-packages/botocore/client.py", line 1021, in _make_api_call
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]:     raise error_class(parsed_response, operation_name)
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: botocore.errorfactory.InvalidSsmlException: An error occurred (InvalidSsmlException) when calling the SynthesizeSpeech operation: Invalid SSML request
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: 2024-04-24 17:40:58.392 - HiveMind-voice-sat - ovos_PHAL_plugin_mk1.arduino:process:63 - INFO - faceplate event: Command: mouth.viseme=3
Apr 24 17:40:58 mark1 hivemind-voice-sat[766]: 2024-04-24 17:40:58.397 - HiveMind-voice-sat - ovos_audio.service:execute_fallback_tts:380 - ERROR - No fallback TTS available and main TTS failed!