Skip to content

Latest commit

 

History

History
256 lines (201 loc) · 10.8 KB

README.md

File metadata and controls

256 lines (201 loc) · 10.8 KB
Raw

sakhi-utility-service

sakhi-utility-service is a service used to support the content service in the Sakhi platform. The service is used to extract context information of chosen attributes from an user's query and to achieve translation of text/audio from one language to another language in text/audio format. The service is built using FastAPI and OpenAI API.

AI-utility

🔧 1. Installation

To use the code, you need to follow these steps:

  1. Clone the repository from GitHub:

    git clone https://github.com/Sunbird-AIAssistant/sakhi-utility-service.git

  2. The code requires Python 3.7 or higher and some additional python packages. To install these packages, run the following command in your terminal:

    pip install requirements.txt

  3. You will need a Cloud Storage account to store the audio file of translated response.

  4. create another file .env which will hold the development credentials and add the following variables. Update the LLM details, Cloud object storage details and Translator details.

    # application environment variables
LOG_LEVEL=<log_level>  # INFO, DEBUG, ERROR
CONFIG_INI_PATH=<config_ini_path> # config.ini

# LLM environment variables
LLM_TYPE=<llm_type> # azure, openai, ollama 
GPT_MODEL=<llm_model_name> # gpt-4
OPENAI_API_KEY=<your_openai_key>

# Translator environment variables
TRANSLATION_TYPE=<translation_type> # bhashini, google, dhruva
BHASHINI_ENDPOINT_URL=<your_bhashini_endpoint_url>
BHASHINI_API_KEY=<your_bhashini_api_key>

# Cloud Storage environment variables
BUCKET_TYPE=<cloud_storage_type> # oci, aws, gcp
BUCKET_ENDPOINT_URL=<cloud_storage_endpoint_url>
BUCKET_REGION_NAME=<cloud_storage_region_name>
BUCKET_NAME=<cloud_storage_bucket_name>
BUCKET_SECRET_ACCESS_KEY=<cloud_storage_secret_access_key>
BUCKET_ACCESS_KEY_ID=<cloud_storage_access_key_id>

# Telemetry environment variables
TELEMETRY_ENDPOINT_URL=<TELEMETRY_ENDPOINT_URL> 
TELEMETRY_LOG_ENABLED=<TELEMETRY_LOG_ENABLED>

🏃🏻 2. Running

Once the above installation steps are completed, run the following command in home directory of the repository in terminal

uvicorn main:app

📃 3. API Specification and Documentation

POST /v1/context

API Function

API is used to extract context information of chosen attributes from an user's query. To achieve the same, Few-shot learning has been implemented which requires a set of 'examples' and necessary 'instructions' to be given to LLM (openAI) in order to generate an answer in the instructed format. Configuration of 'instructions' and 'examples' are available in 'config.ini'.

Supported language codes in request:

en,bn,gu,hi,kn,ml,mr,or,pa,ta,te

Request

Required inputs are 'text', 'audio' and 'language'.

Either of the 'text'(string) or 'audio'(string) should be present. If both the values are given, exception is thrown. Another requirement is that the 'language' should be same as the one given in text and audio (i.e, if you pass English as 'language', then your 'text'/'audio' should contain queries in English language). The audio should either contain a publicly downloadable url of mp3 file or base64 encoded text of the mp3.

{
    "text": "ನನ್ನ ಮಗುವಿಗೆ ವಾಟರ್ ಪೇಂಟಿಂಗ್ ಅನ್ನು ಹೇಗೆ ಕಲಿಸುವುದು",
    "language": "kn"
}

Successful Response

{
    "input": {
        "sourceText": "ನನ್ನ ಮಗುವಿಗೆ ವಾಟರ್ ಪೇಂಟಿಂಗ್ ಅನ್ನು ಹೇಗೆ ಕಲಿಸುವುದು",
        "englishText": "How to Teach My Child Water Painting"
    },
    "context": {
        "category": [
            "Activities"
        ],
        "persona": [
            "Parent"
        ],
        "age": [
            "3-5"
        ],
        "keywords": [
            "water painting"
        ],
        "domain": [
            "Aesthetic and Cultural Development"
        ],
        "curricularGoal": [
            "CG-12: Children develop abilities and sensibilities in visual and performing arts and express their emotions through art in meaningful and joyful ways"
        ]
    }
}

POST /v1/translation

API Function

API is used to achieve translation of text/audio from one language to another language in text/audio format. To achieve the same, Bhashini has been integrated. OCI object storage has been used to store translated audio files when audio is chosen as target output format.

Supported language codess in request:

en,bn,gu,hi,kn,ml,mr,or,pa,ta,te

Request

Required inputs are 'input.text', 'input.audio', 'input.language', 'output.language' and 'output.format'.

Either of the 'text'(string) or 'audio'(string) should be present in the 'input'. If both the values are given, exception is thrown. Another requirement is that the 'input.language' should be same as the one given in text and audio (i.e, if you pass English as 'input.language', then your 'text'/'audio' should contain queries in English language). The audio should either contain a publicly downloadable url of mp3 file or base64 encoded text of the mp3. If 'output.language' is not passed in the request, 'English' will be the default value. 'output.format' can only take 'text'/'audio' as values. If 'output.format' is not passed in the request, 'text' will be the default value.

{
    "input": {
        "text": "How to Teach Kids to Play Games",
        "language": "en"
    },
    "output": {
        "language": "kn",
        "format": "text"
    }
}

Successful Response

{
    "translation": {
        "text": "ಮಕ್ಕಳಿಗೆ ಆಟವಾಡಲು ಕಲಿಸುವುದು ಹೇಗೆ?",
        "audio": null
    }
}

🚀 4. Deployment

This repository comes with a Dockerfile. You can use this dockerfile to deploy your version of this application to Cloud Run. Make the necessary changes to your dockerfile with respect to your new changes. (Note: The given Dockerfile will deploy the base code without any error, provided you added the required environment variables (mentioned in the .env file) to either the Dockerfile or the cloud run revision)

Feature request and contribution

  • We are currently in the alpha stage and hence need all the inputs, feedbacks and contributions we can.
  • Kindly visit our project board to see what is it that we are prioritizing.

UTILITY SERVICE SERVER DEPLOYMENT:

Making System ready for Docker:

sudo apt-get update
sudo apt-get install ca-certificates curl gnupg
sudo install -m 0755 -d /etc/apt/keyrings
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpg
sudo chmod a+r /etc/apt/keyrings/docker.gpg

# Add the repository to Apt sources:
echo \
"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \
$(. /etc/os-release && echo "$VERSION_CODENAME") stable" | \
sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt-get update

Installing latest Docker:

$ sudo apt-get install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

Installing PIP:

$ sudo apt install python3-pip

Clone Repo:

$ git clone https://github.com/DJP-Digital-Jaaduii-Pitara/sakhi-utility-service.git

CD to cloned repo:

$ cd sakhi-utility-service

Build Docker Image of the repo:

$ sudo docker build -t sakhi-utility-image .

Create Container:

$ sudo docker run -d -p 8000:8000 --name sakhi-utility-service \
-e LLM_TYPE=$LLM_TYPE \
-e GPT_MODEL=$GPT_MODEL
-e OPENAI_API_KEY=$OPENAI_API_KEY \
-e LOG_LEVEL=INFO  \
-e CONFIG_INI_PATH=config.ini \
-e TRANSLATION_TYPE=$TRANSLATION_TYPE \
-e BHASHINI_ENDPOINT_URL=$BHASHINI_ENDPOINT_URL \
-e BHASHINI_API_KEY=$BHASHINI_API_KEY \
-e BUCKET_TYPE=$BUCKET_TYPE \
-e BUCKET_ENDPOINT_URL=$BUCKET_ENDPOINT_URL \
-e BUCKET_REGION_NAME=$BUCKET_REGION_NAME \
-e BUCKET_NAME=$BUCKET_NAME \
-e BUCKET_SECRET_ACCESS_KEY=$BUCKET_SECRET_ACCESS_KEY \
-e BUCKET_ACCESS_KEY_ID=$BUCKET_ACCESS_KEY_ID \
-e TELEMETRY_ENDPOINT_URL=$TELEMETRY_ENDPOINT_URL \
-e TELEMETRY_LOG_ENABLED=$TELEMETRY_LOG_ENABLED \
sakhi-utility-image

Configuration (config.ini)

Variable Description Default Value
few_shot_config.instructions System prompt for GEN AI to perform the context recognition for user's query
few_shot_config.examples Context examples inserted to System prompt for GEN AI to perform few shot learning
lang_code.supported_lang_codes Supported languages by the service en,bn,gu,hi,kn,ml,mr,or,pa,ta,te
min_words.length Minimum length of words in user's query for which context extraction get enabled by Gen AI 6
telemetry.telemetry_log_enabled Flag to enable or disable telemetry events logging to Sunbird Telemetry service true
telemetry.environment service environment from where telemetry is generated from, in telemetry service dev
telemetry.service_id service identifier to be passed to Sunbird telemetry service
telemetry.service_ver service version to be passed to Sunbird telemetry service
telemetry.actor_id service actor id to be passed to Sunbird telemetry service
telemetry.channel channel value to be passed to Sunbird telemetry service
telemetry.pdata_id pdata_id value to be passed to Sunbird telemetry service
telemetry.events_threshold telemetry events batch size upon which events will be passed to Sunbird telemetry service 5