New runner. Manual squash. (#2110)

* New runner. Manual squash.

* Network host.

* Put back trufflehog with proper extension.

* No network host ?

* Moving buildx install after tailscale ?

* 1.79

.github/workflows/ci_build.yaml

@@ -0,0 +1,36 @@
+name: CI build
+
+on:
+  push:
+    branches:
+      - 'main'
+    tags:
+      - 'v*'
+  pull_request:
+    paths:
+      - ".github/workflows/build.yaml"
+      - "integration-tests/**"
+      - "server/**"
+      - "proto/**"
+      - "router/**"
+      - "launcher/**"
+      - "Cargo.lock"
+      - "rust-toolchain.toml"
+      - "Dockerfile"
+      - "Dockerfile_amd"
+      - "Dockerfile_intel"
+    branches:
+      - 'main'
+
+jobs:
+  build:
+    strategy:
+      # super important if you want to see all results, even if one fails
+      # fail-fast is true by default
+      fail-fast: false
+      matrix:
+        hardware: ["cuda", "rocm", "intel"]
+    uses: ./.github/workflows/build.yaml # calls the one above ^
+    with:
+      hardware: ${{ matrix.hardware }}
+    secrets: inherit

.github/workflows/integration_tests.yaml

@@ -0,0 +1,41 @@
+name: Integration tests
+
+on:
+  workflow_call:
+    inputs:
+      docker_image:
+        type: string
+        description: Hardware
+        required: true
+      docker_devices:
+        type: string
+        description: Hardware
+      runs_on:
+        type: string
+        required: true
+        description: Hardware to run integration tests
+jobs:
+  integration_tests:
+    concurrency:
+      group: ${{ github.workflow }}-${{ github.job }}-${{ github.head_ref || github.run_id }}
+      cancel-in-progress: true
+    runs-on: ${{ inputs.runs_on }}
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+      - name: Inject slug/short variables
+        uses: rlespinasse/github-slug-action@v4.4.1
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.9
+      - name: Install
+        run: |
+          make install-integration-tests
+      - name: Run tests
+        run: |
+          export DOCKER_VOLUME=/mnt/cache
+          export DOCKER_IMAGE=${{ inputs.docker_image }}
+          export DOCKER_DEVICES=${{ inputs.docker_devices }}
+          export HUGGING_FACE_HUB_TOKEN=${{ secrets.HUGGING_FACE_HUB_TOKEN }}
+          pytest -s -vv integration-tests

integration-tests/conftest.py

@@ -34,6 +34,7 @@ from text_generation.types import (
 DOCKER_IMAGE = os.getenv("DOCKER_IMAGE", None)
 HUGGING_FACE_HUB_TOKEN = os.getenv("HUGGING_FACE_HUB_TOKEN", None)
 DOCKER_VOLUME = os.getenv("DOCKER_VOLUME", "/data")
+DOCKER_DEVICES = os.getenv("DOCKER_DEVICES")
 
 
 class ResponseComparator(JSONSnapshotExtension):
@@ -453,6 +454,18 @@ def launcher(event_loop):
         if DOCKER_VOLUME:
             volumes = [f"{DOCKER_VOLUME}:/data"]
 
+        if DOCKER_DEVICES:
+            devices = DOCKER_DEVICES.split(",")
+            visible = os.getenv("ROCR_VISIBLE_DEVICES")
+            if visible:
+                env["ROCR_VISIBLE_DEVICES"] = visible
+            device_requests = []
+        else:
+            devices = []
+            device_requests = [
+                docker.types.DeviceRequest(count=gpu_count, capabilities=[["gpu"]])
+            ]
+
         container = client.containers.run(
             DOCKER_IMAGE,
             command=args,
@@ -460,9 +473,8 @@ def launcher(event_loop):
             environment=env,
             auto_remove=False,
             detach=True,
-            device_requests=[
-                docker.types.DeviceRequest(count=gpu_count, capabilities=[["gpu"]])
-            ],
+            device_requests=device_requests,
+            devices=devices,
             volumes=volumes,
             ports={"80/tcp": port},
             shm_size="1G",