db-copilot-tool 0.0.5

nl2sql tool for prompt flow

nl2sql tool for prompt flow

How to create promptflow endpoint your workspace

1. Prepare your workspace

Open your workspace in Azure ML studio portal
Example Studio Portal URL:
https://main.ml.azure.com/?wsid=/subscriptions/1b75927d-563b-49d2-bf8a-772f7d6a170e/resourceGroups/ragdev/providers/Microsoft.MachineLearningServices/workspaces/RAGDev&flight=promptDesigner&tid=72f988bf-86f1-41af-91ab-2d7cd011db47

2. Create a ci (compute instance)

Create a ci with name with prefix "ci-lin-cpu-0-" , for example ci-lin-cpu-0-alan

3. Config Customer Application on your ci

3.1 Application Name

Application Name has to be: promptflow-runtime

3.2 Port

Target Port and Published Port have to be: 8080

3.3 Docker Image

docker image can be modulesdkpreview/prt:tag.
For example: modulesdkpreview/prt:v20230420-140738

3.4 Restart ci

4. Install dbcopilot on your ci

4.1 Open Terminal of your ci

4.2 upload whl files to your ci

dbcopilot-0.2.0-py3-none-any.whl (get it from https://dev.azure.com/TScience/NL2Code/_artifacts/feed/AIMS.TScience.NL2Code/PyPI/dbcopilot/overview/0.3.0)
db_copilot_tool-0.1.0-py3-none-any.whl
embeddingstore-0.0.1-py3-none-any.whl (optional)
promptflow_sdk-0.0.1-py3-none-any.whl (optional)

4.3 Install whl files in promptflow container

Find your promptflow container id with command: docker ps -a
Copy whl files to promptflow container with command: docker cp xxx.whl <container_id>:/xxx.whl
Login promptflow container with command: docker exec -it <container_id> /bin/bash
install whl files with command: pip install xxx.whl

4.4 Install Microsoft ODBC in promptflow container

Reference Doc: https://learn.microsoft.com/en-us/sql/connect/odbc/linux-mac/installing-the-microsoft-odbc-driver-for-sql-server?view=sql-server-ver16&tabs=ubuntu18-install%2Calpine17-install%2Cdebian8-install%2Credhat7-13-install%2Crhel7-offline#18
Commands:
apt-get update
apt-get install -y curl lsb-core
curl https://packages.microsoft.com/keys/microsoft.asc | apt-key add -
curl https://packages.microsoft.com/config/ubuntu/$(lsb_release -rs)/prod.list > /etc/apt/sources.list.d/mssql-release.list
apt-get update
ACCEPT_EULA=Y apt-get install -y msodbcsql18
apt-get install -y unixodbc-dev

4.5 Restart promptflow container

exit docker stop container_id
docker start container_id

5. Copy Local DB file to promptflow container (optional)

6. Get promptflow endpoint url

promptflow endpoint url can be found in the ci overview page, click "promptflow-runtime" link