Add e2e test workflow

This workflow runs an end-to-end test workflow using `ilab` based on
the latest instructlab-sdg code. This is based on the e2e workflow in
the instructlab repo.

It currently runs against whatever is in `main` for `instructlab`.

Signed-off-by: Russell Bryant <[email protected]>

.github/actionlint.yaml

@@ -0,0 +1,3 @@
+self-hosted-runner:
+  labels:
+  - ubuntu-gpu

.github/workflows/e2e.yml

@@ -0,0 +1,92 @@
+# SPDX-License-Identifier: Apache-2.0
+
+name: E2E test
+
+on:
+  push:
+    branches:
+      - "main"
+      - "release-**"
+    paths:
+      - '**.py'
+      - 'pyproject.toml'
+      - 'requirements*.txt'
+      - '.github/workflows/e2e.yml'
+  pull_request:
+    branches:
+      - "main"
+      - "release-**"
+    paths:
+      - '**.py'
+      - 'pyproject.toml'
+      - 'requirements*.txt'
+      - '.github/workflows/e2e.yml'
+
+jobs:
+  e2e:
+    runs-on: ubuntu-gpu
+
+    permissions:
+      pull-requests: write
+
+    steps:
+      # No step-security/harden-runner since this is a self-hosted runner
+      - name: Checkout
+        uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
+        with:
+          # https://github.com/actions/checkout/issues/249
+          fetch-depth: 0
+
+      - name: Install Packages
+        run: |
+          sudo apt-get install -y cuda-toolkit git cmake build-essential virtualenv
+          nvidia-smi
+          ls -l /dev/nvidia*
+
+      - name: Setup Python 3.11
+        uses: actions/setup-python@82c7e631bb3cdc910f68e0081d67478d79c6982d # v5.1.0
+        with:
+          python-version: 3.11
+          cache: pip
+          cache-dependency-path: |
+            **/pyproject.toml
+            **/requirements*.txt
+
+      - name: Remove llama-cpp-python from cache
+        run: |
+          pip cache remove llama_cpp_python
+
+      - name: Cache huggingface
+        uses: actions/cache@0c45773b623bea8c8e75f6c82b208c3cf94ea4f9 # v4.0.2
+        with:
+          path: ~/.cache/huggingface
+          # config contains DEFAULT_MODEL
+          key: huggingface-${{ hashFiles('src/instructlab/configuration.py') }}
+
+      - name: Install instructlab and instructlab-sdg
+        run: |
+          export PATH="/home/runner/.local/bin:/usr/local/cuda/bin:$PATH"
+          python3 -m venv venv
+          . venv/bin/activate
+          git clone https://github.com/instructlab/instructlab.git
+          cd instructlab
+          sed 's/\[.*\]//' requirements.txt > constraints.txt
+          python3 -m pip cache remove llama_cpp_python
+          CMAKE_ARGS="-DLLAMA_CUBLAS=on" python3 -m pip install --no-binary llama_cpp_python -c constraints.txt llama_cpp_python
+          # needed for --4-bit-quant option to ilab train
+          python3 -m pip install bitsandbytes
+          # install instructlab
+          python3 -m pip install .
+          cd ..
+          # Install instructlab-sdg
+          python3 -m pip install .
+
+      - name: Run e2e test
+        run: |
+          . venv/bin/activate
+          ./instructlab/scripts/basic-workflow-tests.sh -cm
+
+      - name: Remove llama-cpp-python from cache
+        if: always()
+        run: |
+          pip cache remove llama_cpp_python