Care Scribe is a plugin for care to add voice auto fill support using external services like OpenAI whisper and Google Speech to Text.
- Voice auto fill support for care
- Support for OpenAI whisper and Google Speech to Text
https://care-be-docs.ohc.network/pluggable-apps/configuration.html
https://github.com/ohcnetwork/care/blob/develop/plug_config.py
To install care scribe, you can add the plugin config in care/plug_config.py as follows:
...
scribe_plug = Plug(
name="care_scribe",
package_name="git+https://github.com/ohcnetwork/care_scribe.git",
version="@master",
configs={
"TRANSCRIBE_SERVICE_PROVIDER_API_KEY": "secret",
"API_PROVIDER": "openai", # or "azure"
"AZURE_API_VERSION": "", # required if API_PROVIDER is "azure"
"AZURE_ENDPOINT": "", # required if API_PROVIDER is "azure"
"AUDIO_MODEL_NAME": "", # model name for OpenAI or custom deployment name for Azure
"CHAT_MODEL_NAME": "", # model name for OpenAI or custom deployment name for Azure
},
)
plugs = [scribe_plug]
...
The following configurations variables are available for Care Scribe:
TRANSCRIBE_SERVICE_PROVIDER_API_KEY
: API key for the transcribe service provider (OpenAI whisper or Google Speech to Text)API_PROVIDER
: The API provider to use for transcription. Can be either "openai" or "azure".AZURE_API_VERSION
: The version of the Azure API to use. This is required ifAPI_PROVIDER
is set to "azure".AZURE_ENDPOINT
: The endpoint for the Azure API. This is required ifAPI_PROVIDER
is set to "azure".AUDIO_MODEL_NAME
: The model name for OpenAI or the custom deployment name for Azure.CHAT_MODEL_NAME
: The model name for OpenAI or the custom deployment name for Azure.
The plugin will try to find the API key from the config first and then from the environment variable.
This project is licensed under the terms of the MIT license.
This plugin was created with Cookiecutter using the ohcnetwork/care-plugin-cookiecutter.