Update agent-speech-processing.yaml files
This commit is contained in:
pabloFuente
parent
1fd49f308c
commit
7c22e68ab5
@ -7,7 +7,7 @@ enabled: false
|
||||
# Maximum CPU load threshold for the agent to accept new jobs. Value between 0 and 1.
|
||||
load_threshold: 1.0
|
||||
|
||||
# Log level for the agent [DEBUG, INFO, WARNING, ERROR, CRITICAL]
|
||||
# Log level for the agent [DEBUG, INFO, WARN, ERROR, CRITICAL]
|
||||
log_level: INFO
|
||||
|
||||
live_captions:
|
||||
@ -16,7 +16,7 @@ live_captions:
|
||||
# - manual: the agent will connect to new Rooms only when your application dictates it by using the Agent Dispatch API.
|
||||
processing: automatic
|
||||
|
||||
# Which speech-to-text AI provider to use [aws, azure, google, openai, azure_openai, groq, deepgram, assemblyai, fal, clova, speechmatics, gladia, sarvam, mistralai, cartesia, soniox, nvidia, vosk]
|
||||
# Which speech-to-text AI provider to use [aws, azure, google, openai, azure_openai, groq, deepgram, assemblyai, fal, clova, speechmatics, gladia, sarvam, mistralai, cartesia, soniox, nvidia, elevenlabs, simplismart, vosk, sherpa]
|
||||
# The custom configuration for the selected provider must be set below
|
||||
provider:
|
||||
|
||||
@ -348,6 +348,41 @@ live_captions:
|
||||
# Language short code for the generated speech. For supported values, see https://docs.spitch.app/concepts/languages
|
||||
language:
|
||||
|
||||
elevenlabs:
|
||||
# API key for ElevenLabs. See https://elevenlabs.io/app/settings/api-keys
|
||||
api_key:
|
||||
# The ElevenLabs STT model to use. Valid values are ["scribe_v1", "scribe_v2", "scribe_v2_realtime"]. See https://elevenlabs.io/docs/overview/models#models-overview
|
||||
model_id:
|
||||
# An ISO-639-1 or ISO-639-3 language_code corresponding to the language of the audio file. Can sometimes improve transcription performance if known beforehand. Defaults to null, in this case the language is predicted automatically
|
||||
language_code:
|
||||
# Custom base URL for the API. Optional.
|
||||
base_url:
|
||||
# Audio sample rate in Hz. Default is 16000.
|
||||
sample_rate:
|
||||
# Whether to tag audio events like (laughter), (footsteps), etc. in the transcription. Only supported for Scribe v1 model. Default is True
|
||||
tag_audio_events:
|
||||
# Whether to include word-level timestamps in the transcription. Default is false.
|
||||
include_timestamps:
|
||||
|
||||
simplismart:
|
||||
# API key for SimpliSmart. See https://docs.simplismart.ai/model-suite/settings/api-keys
|
||||
api_key:
|
||||
# Model identifier for the backend STT model. One of ["openai/whisper-large-v2", "openai/whisper-large-v3", "openai/whisper-large-v3-turbo"]
|
||||
# Default is "openai/whisper-large-v3-turbo"
|
||||
model:
|
||||
# Language code for transcription (default: "en"). See https://docs.simplismart.ai/get-started/playground/transcription-models#supported-languages-with-their-codes
|
||||
language:
|
||||
# Operation to perform. "transcribe" converts speech to text in the original language, "translate" translates into English. Default is "transcribe".
|
||||
task:
|
||||
# If true, disables timestamp generation in transcripts. Default is true
|
||||
without_timestamps:
|
||||
# Minimum duration (ms) for a valid speech segment. Default is 0
|
||||
min_speech_duration_ms:
|
||||
# Decoding temperature (affects randomness). Default is 0.0
|
||||
temperature:
|
||||
# Whether to permit multilingual recognition. Default is false
|
||||
multilingual:
|
||||
|
||||
vosk:
|
||||
# Vosk language model. This provider requires docker_image "docker.io/openvidu/agent-speech-processing-vosk"
|
||||
# Below is the list of pre-installed models in the container (available at https://alphacephei.com/vosk/models):
|
||||
|
||||
@ -7,7 +7,7 @@ enabled: false
|
||||
# Maximum CPU load threshold for the agent to accept new jobs. Value between 0 and 1.
|
||||
load_threshold: 1.0
|
||||
|
||||
# Log level for the agent [DEBUG, INFO, WARNING, ERROR, CRITICAL]
|
||||
# Log level for the agent [DEBUG, INFO, WARN, ERROR, CRITICAL]
|
||||
log_level: INFO
|
||||
|
||||
live_captions:
|
||||
@ -16,7 +16,7 @@ live_captions:
|
||||
# - manual: the agent will connect to new Rooms only when your application dictates it by using the Agent Dispatch API.
|
||||
processing: automatic
|
||||
|
||||
# Which speech-to-text AI provider to use [aws, azure, google, openai, azure_openai, groq, deepgram, assemblyai, fal, clova, speechmatics, gladia, sarvam, mistralai, cartesia, soniox, nvidia, vosk]
|
||||
# Which speech-to-text AI provider to use [aws, azure, google, openai, azure_openai, groq, deepgram, assemblyai, fal, clova, speechmatics, gladia, sarvam, mistralai, cartesia, soniox, nvidia, elevenlabs, simplismart, vosk, sherpa]
|
||||
# The custom configuration for the selected provider must be set below
|
||||
provider:
|
||||
|
||||
@ -348,6 +348,41 @@ live_captions:
|
||||
# Language short code for the generated speech. For supported values, see https://docs.spitch.app/concepts/languages
|
||||
language:
|
||||
|
||||
elevenlabs:
|
||||
# API key for ElevenLabs. See https://elevenlabs.io/app/settings/api-keys
|
||||
api_key:
|
||||
# The ElevenLabs STT model to use. Valid values are ["scribe_v1", "scribe_v2", "scribe_v2_realtime"]. See https://elevenlabs.io/docs/overview/models#models-overview
|
||||
model_id:
|
||||
# An ISO-639-1 or ISO-639-3 language_code corresponding to the language of the audio file. Can sometimes improve transcription performance if known beforehand. Defaults to null, in this case the language is predicted automatically
|
||||
language_code:
|
||||
# Custom base URL for the API. Optional.
|
||||
base_url:
|
||||
# Audio sample rate in Hz. Default is 16000.
|
||||
sample_rate:
|
||||
# Whether to tag audio events like (laughter), (footsteps), etc. in the transcription. Only supported for Scribe v1 model. Default is True
|
||||
tag_audio_events:
|
||||
# Whether to include word-level timestamps in the transcription. Default is false.
|
||||
include_timestamps:
|
||||
|
||||
simplismart:
|
||||
# API key for SimpliSmart. See https://docs.simplismart.ai/model-suite/settings/api-keys
|
||||
api_key:
|
||||
# Model identifier for the backend STT model. One of ["openai/whisper-large-v2", "openai/whisper-large-v3", "openai/whisper-large-v3-turbo"]
|
||||
# Default is "openai/whisper-large-v3-turbo"
|
||||
model:
|
||||
# Language code for transcription (default: "en"). See https://docs.simplismart.ai/get-started/playground/transcription-models#supported-languages-with-their-codes
|
||||
language:
|
||||
# Operation to perform. "transcribe" converts speech to text in the original language, "translate" translates into English. Default is "transcribe".
|
||||
task:
|
||||
# If true, disables timestamp generation in transcripts. Default is true
|
||||
without_timestamps:
|
||||
# Minimum duration (ms) for a valid speech segment. Default is 0
|
||||
min_speech_duration_ms:
|
||||
# Decoding temperature (affects randomness). Default is 0.0
|
||||
temperature:
|
||||
# Whether to permit multilingual recognition. Default is false
|
||||
multilingual:
|
||||
|
||||
vosk:
|
||||
# Vosk language model. This provider requires docker_image "docker.io/openvidu/agent-speech-processing-vosk"
|
||||
# Below is the list of pre-installed models in the container (available at https://alphacephei.com/vosk/models):
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user