Adds an action for detecting a regression in partion

.github/workflows/partition-benchmark.yaml

@@ -0,0 +1,120 @@
+name: Partition Benchmark
+
+# Runs on every PR targeting main to detect regressions.
+# Can also be triggered manually to establish or inspect a new baseline.
+on:
+  pull_request:
+    branches: [main]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+env:
+  NLTK_DATA: ${{ github.workspace }}/nltk_data
+  PYTHON_VERSION: "3.12"
+  # Number of times to run the full benchmark suite.
+  NUM_ITERATIONS: "3"
+  # 20% threshold for now and tune later
+  REGRESSION_THRESHOLD: "0.20"
+  # Increment to change cache key when benchmark-affecting dependencies are updated, to ensure clean slate runs.
+  CACHE_VERSION: "v2"
+  # S3 location for metrics – matches core-product convention.
+  S3_METRICS_BUCKET_KEY: utic-metrics/ci-metrics
+  S3_BENCHMARK_PATH: open-source/partition-benchmark/benchmark_best.json
+
+jobs:
+  setup:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: ./.github/actions/base-cache
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+  benchmark:
+    name: Measure and compare partition() runtime
+    runs-on: ubuntu-latest
+    needs: [setup]
+
+    steps:
+
+      - uses: actions/checkout@v4
+
+      - uses: ./.github/actions/base-cache
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Install system dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y libmagic-dev poppler-utils libreoffice
+          sudo add-apt-repository -y ppa:alex-p/tesseract-ocr5
+          sudo apt-get update
+          sudo apt-get install -y tesseract-ocr tesseract-ocr-kor
+
+
+      - name: Restore HuggingFace model cache
+        uses: actions/cache/restore@v4
+        with:
+          path: ~/.cache/huggingface
+          key: hf-models-${{ runner.os }}-${{ env.CACHE_VERSION }}-${{ github.sha }}
+          restore-keys: |
+            hf-models-${{ runner.os }}-${{ env.CACHE_VERSION }}-
+            hf-models-${{ runner.os }}-
+
+
+      - name: Run partition benchmark
+        env:
+          NUM_ITERATIONS: ${{ env.NUM_ITERATIONS }}
+        run: |
+          uv run --no-sync python scripts/performance/benchmark_partition.py \
+            benchmark_results.json
+
+      - name: Save HuggingFace model cache
+        uses: actions/cache/save@v4
+        with:
+          path: ~/.cache/huggingface
+          key: hf-models-${{ runner.os }}-${{ env.CACHE_VERSION }}-${{ github.sha }}
+
+
+      - name: Download previous best from S3
+        continue-on-error: true
+        env:
+          AWS_ACCESS_KEY_ID: ${{ secrets.S3_EVAL_ACCESS_KEY }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.S3_EVAL_SECRET_KEY }}
+        run: |
+          aws s3 cp \
+            "s3://${{ env.S3_METRICS_BUCKET_KEY }}/${{ env.S3_BENCHMARK_PATH }}" \
+            benchmark_best.json
+
+
+      - name: Compare results against stored best
+        id: compare
+        run: |
+          uv run --no-sync python scripts/performance/compare_benchmark.py \
+            benchmark_results.json \
+            benchmark_best.json \
+            ${{ env.REGRESSION_THRESHOLD }}
+
+
+      - name: Upload best result to S3
+        continue-on-error: true
+        env:
+          AWS_ACCESS_KEY_ID: ${{ secrets.S3_EVAL_ACCESS_KEY }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.S3_EVAL_SECRET_KEY }}
+        run: |
+          aws s3 cp \
+            benchmark_best.json \
+            "s3://${{ env.S3_METRICS_BUCKET_KEY }}/${{ env.S3_BENCHMARK_PATH }}"
+
+
+      - name: Upload benchmark artifacts
+        if: always()
+        uses: actions/upload-artifact@v4
+        with:
+          name: benchmark-results-${{ github.sha }}
+          path: |
+            benchmark_results.json
+            benchmark_best.json
+          retention-days: 30

.gitignore

@@ -209,4 +209,15 @@ annotated/
 pcaps
 python-output
 
-.vs/
+.vs/
+# Partition benchmark generated output
+benchmark_results.json
+scripts/performance/partition-speed-test/benchmark_results.json
+
+# Partition benchmark generated output
+benchmark_results.json
+scripts/performance/partition-speed-test/benchmark_results.json
+
+# Partition benchmark generated output
+benchmark_results.json
+scripts/performance/partition-speed-test/benchmark_results.json

CHANGELOG.md

@@ -1,3 +1,6 @@
+## 0.21.4
+- Add a git action to test regression in runtime performance
+
 ## 0.21.3
 
 ### Enhancements
@@ -17,6 +20,15 @@
 ### Fixes
 - **Replace NLTK with spaCy to remediate CVE-2025-14009**: NLTK's downloader uses `zipfile.extractall()` without path validation, enabling RCE via malicious packages (CVSS 10.0, no patch available). spaCy models install as pip packages, eliminating the vulnerable downloader entirely.
 
+||||||| parent of f7a4a80e (fix: update depdencies (#4247))
+=======
+>>>>>>> f7a4a80e (fix: update depdencies (#4247))
+||||||| parent of 0758132c (bump version)
+=======
+## 0.20.9
+- add an action to test the partition seepd in high_res and fast
+
+>>>>>>> 0758132c (bump version)
 ## 0.20.8
 
 ### Fixes