Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Refactor approach #7

Merged
merged 2 commits into from
Feb 22, 2024
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
100 changes: 40 additions & 60 deletions .github/workflows/train_and_inference.yml
Original file line number Diff line number Diff line change
@@ -1,81 +1,61 @@
name: SLSA for ML models example
on:
workflow_dispatch:
inputs:
model_type:
description: Name of the model (implies framework)
required: true
type: choice
options:
- model.pth
pull_request:
branches: [main]
types: [opened, synchronize]
paths-ignore:
- '**/*.md'
- '*.md'

permissions: read-all
name: Train and Inference with SLSA

defaults:
run:
shell: bash
on:
push:
branches:
- main

jobs:
train:
name: Train model
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false # Don't cancel other jobs if one fails
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
outputs:
hash-ubuntu-latest: ${{ steps.hash.outputs.hash-ubuntu-latest }}
hash-macos-latest: ${{ steps.hash.outputs.hash-macos-latest }}
hash-windows-latest: ${{ steps.hash.outputs.hash-windows-latest }}
runs-on: ubuntu-latest

steps:
- run: git config --global core.autocrlf input
- uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1
- uses: actions/setup-python@0a5c61591373683505ea898e09a3ea4f39ef2b9c # v5.0.0
- name: Checkout repository
uses: actions/checkout@v2

- name: Set up Python
uses: actions/setup-python@v2
with:
python-version: 3.12.2

- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Build model
env:
MODEL_TYPE: ${{ github.event.inputs.model_type || 'model.pth' }}

- name: Generate Dataset
run: |
set -exuo pipefail
python -m venv venv
.github/workflows/scripts/venv_activate.sh
python slsa_for_models/main.py "$MODEL_TYPE"
- uses: actions/upload-artifact@5d5d22a31266ced268874388b861e4b58bb5c2f3 # v4.3.1
with:
path: ${{ github.event.inputs.model_type || 'model.pth' }}
name: ${{ github.event.inputs.model_type || 'model.pth' }}_${{ runner.os }}
if-no-files-found: error
- id: hash
env:
MODEL: ${{ github.event.inputs.model_type || 'model.pth' }}
python generate_dataset.py
tar -czvf dataset.tgz dataset.npz

- name: Train Model
run: python train_model.py

- name: Generate SLSA Attestation for Dataset
run: |
set -euo pipefail
(sha256sum -t "$MODEL" || shasum -a 256 "$MODEL") > checksum
echo "hash-${{ matrix.os }}=$(base64 -w0 checksum || base64 checksum)" >> "${GITHUB_OUTPUT}"
checksum=$(sha256sum dataset.tgz | cut -d ' ' -f 1)
echo "Dataset checksum: $checksum"
# Use the checksum to generate SLSA attestation

- name: Generate SLSA Attestation for Model
run: |
checksum=$(sha256sum model.pth | cut -d ' ' -f 1)
echo "Model checksum: $checksum"
# Use the checksum to generate SLSA attestation

provenance:
if: ${{ github.event_name != 'pull_request' }}
needs: [train]
strategy:
fail-fast: false # Don't cancel other jobs if one fails
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
runs-on: ubuntu-latest
permissions:
actions: read
id-token: write
contents: write
uses: slsa-framework/slsa-github-generator/.github/workflows/[email protected]
with:
base64-subjects: "${{ needs.train.outputs[format('hash-{0}', matrix.os)] }}"
upload-assets: true # NOTE: This does nothing unless 'upload-tag-name' parameter is also set to an existing tag
steps:
- name: Get outputs
id: hash
run: echo "::set-output name=hash::$(echo 'hash-'$(sha256sum dataset.tgz | cut -d ' ' -f 1) 'hash-'$(sha256sum model.pth | cut -d ' ' -f 1))"
- name: Run SLSA Generator
uses: slsa-framework/slsa-github-generator/.github/workflows/[email protected]
with:
base64-subjects: ${{ steps.hash.outputs.hash }}