Merge pull request #10 from ashiq-km/yml_workflow

Yml workflow

Author: ashiq-km

.dvc/config

@@ -0,0 +1,7 @@
+[core]
+    remote = storage
+['remote "storage"']
+    url = gdrive://1gRkn11DXoANBF9CGXOsiZyJCjLKoeO_4
+    gdrive_client_id = 49626850770-j19rjs1gp2p67qam8q7dnfoeq3oq1475.apps.googleusercontent.com
+    gdrive_client_secret = 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
+    gdrive_acknowledge_abuse = true

.github/workflows/model-training-test.yml

@@ -0,0 +1,62 @@
+name: Model Training & Testing
+
+
+on:
+  push:
+    branches:  [main, yml_workflow]
+  pull_request:
+    branches:  [main]
+
+  workflow_dispatch: # Allows manual trigger
+
+jobs:
+  train-and-test:
+    runs-on: ubuntu-latest
+
+    env:
+      PYTHONPATH: ${{ github.workspace }}
+
+    steps:
+      - name: checkout code
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-verson: '3.10'
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements.txt
+          pip install "dvc[gdrive]"
+
+
+      - name: Configure DVC with Service Account
+        env:
+          GCP_SA_KEY: ${{ secrets.GCP_SERVICE_ACCOUNT_KEY }}
+        run: |
+          echo "$GCP_SA_KEY" > service_account.json
+          dvc remote modify storage gdrive_use_service_account true
+          dvc remote modify storage gdrive_service_account_json_file_path service_account.json
+        working-directory: .
+
+
+      - name: Pull data from Google drive
+        run: |
+          dvc pull
+
+      - name: Run Pre-processing
+        run: |
+          python src/preprocess.py
+
+      - name: Train Model
+        run: |
+         python src/train.py
+
+      - name: Upload model artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: trained-model
+          path: data/models/
+          retention-days: 7

.gitignore

@@ -18,7 +18,11 @@ data/models/
 # Tools
 
 mlruns/
-.dvc/
 .env
 
-*.html
+*.html
+
+# DVC - ignore cache and temp, but track config
+.dvc/cache/
+.dvc/tmp/
+.dvc/config.local