.github/workflows/ci.yml

name: Update data and Update model with Hugging Face

on:
  # schedule:
  #   - cron: '0 0 * * *' # This cron expression runs the job daily
  # workflow_dispatch:
  push:
    branches:
      - main

jobs:
  crawl:
    runs-on: ubuntu-latest

    steps:
      - name: Checkout
        uses: actions/checkout@v4.1.6
        with:
          ref: main

      - name: Setup Python
        uses: actions/setup-python@v5.1.0
        with:
          python-version: 3.11

      - name: Install Dependencies
        run: |
          python -m pip install --upgrade pip 
          if [ -f requirements.txt ]; then pip install -r requirements.txt; fi

      - name: Run update
        run: python ./main/main_update.py

      - name: Upload output
        uses: actions/upload-artifact@v4
        with:
          name: movies_data.csv
          path: merge_data/movies_data.csv

  update_model:
    runs-on: ubuntu-latest
    needs: crawl

    steps:
      - name: Checkout
        uses: actions/checkout@v4.1.6
        with:
          ref: main

      - name: Setup Python
        uses: actions/setup-python@v5.1.0
        with:
          python-version: 3.11
          
      - name: Install Dependencies
        run: |
          python -m pip install --upgrade pip 
          pip install transformers datasets huggingface_hub
          if [ -f requirements.txt ]; then pip install -r requirements.txt; fi

      - name: Download artifact
        uses: actions/download-artifact@v4
        with:
          name: movies_data.csv
          path: merge_data/

      - name: Update model using Hugging Face
        run: |
          # Download a pre-trained model from Hugging Face
          from transformers import AutoModelForSequenceClassification, AutoTokenizer

          model_name = "distilbert-base-uncased"
          tokenizer = AutoTokenizer.from_pretrained(model_name)
          model = AutoModelForSequenceClassification.from_pretrained(model_name)

          # Your custom model training or updating logic
          python update_model.py

          # Save updated model to Hugging Face Hub (Optional)
          from huggingface_hub import HfApi, Repository
          repo = Repository(local_dir="./model_output", clone_from="your-hf-username/your-model-name")
          repo.push_to_hub(commit_message="Updated model")

      - name: Commit and push artifacts
        env: 
          GH_PAT: ${{ secrets.GH_PAT }}
        run: |
          git config --global user.name 'github-actions[bot]'
          git config --global user.email '41898282+github-actions[bot]@users.noreply.github.com'
          git add model_efa/*.pkl
          git add merge_data/final_merged.csv
          git commit -m 'Update .pkl files in model_efa directory'
          git push https://x-access-token:${{ secrets.GH_PAT }}@github.com/QuanTH02/2023.2-Data-Science.git main
        continue-on-error: true
  
  # pull_main:
  #   runs-on: [self-hosted, quan_win]
  #   needs: [crawl, update_model]

  #   steps:
  #     - name: Checkout code
  #       uses: actions/checkout@v4.1.6
  #       with:
  #         ref: main

  #     - name: Pull latest changes
  #       run: |
  #         git pull origin main