TomFern/semaphore-demo-pipelinegpt

Demo project to create an enriched ChatGPT agent using embeddings

PipelineGPT: Generate Semaphore Pipelines with OpenAI

Demo project to build an enriched ChatGPT responses using embeddings.

Requirements

• Python 3
• A OpenAI API account (paid)
• A Starter Plan Pinecone account (free)

Installation

1. Clone the repository
2. Create a virtualenv for Python:

   $ virtualenv venv
   $ source venv/bin/activate

3. Install dependencies

   $ pip install -r requirements.txt

4. Initialize environment file. Add your Pinecone and OpenAI API Keys

   $ cp env-example .env
   $ nano .env

Create embeddings database

1. Clone the Semaphore docs repository

   $ git clone https://github.com/semaphoreci/docs.git

2. Create a Pinecone database

   $ source venv/bin/activate
   $ source .env
   $ cd database
   $ python db_create.py

3. Create embeddings and upload them to Pinecone.

   $ source venv/bin/activate
   $ source .env
   $ cd database
   $ python index_docs.py /path/repository/semaphoreci/docs
   
   Extracting YAML from Markdown files in /Users/tom/r/docs
   100%|████████████████████████| 164/164 [00:00<00:00, 8206.95it/s]
   Creating embeddings and uploading vectors to database
   100%|████████████████████████| 4/4 [00:09<00:00,  2.27s/it]
   Database contains 79 vectors.

Run your query

$ source venv/bin/activate
$ source .env
$ python query.py "Create a CI pipeline to build and upload a Docker image to Docker Hub"
Found 22 contexts for your query
Working on your query...

Answer:

version: v1.0
name: Docker Build and Push
agent:
  machine:
    type: e1-standard-2
    os_image: ubuntu1804

blocks:
  - name: "Build and Push Docker Image"
    task:
      jobs:
        - name: "Build and Push"
          commands:
            - checkout
            - docker build -t <repository>/<image>:<tag> .
            - echo "${DOCKER_PASSWORD}" | docker login -u "${DOCKER_USERNAME}" --password-stdin
            - docker push <repository>/<image>:<tag>
      secrets:
        - name: dockerhub
promotions:
  - name: Deploy to Kubernetes
    pipeline_file: deploy-k8s.yml
    auto_promote:
      when: "result = 'passed'"