Use ORT nightly in CI

.github/workflows/gpu_l4_golden_parity.yml

@@ -43,14 +43,6 @@ jobs:
         with:
           python-version: "3.12"
 
-      - name: Cache pip packages
-        uses: actions/cache@v5
-        with:
-          path: ~/.cache/pip
-          key: pip-gpu-${{ hashFiles('pyproject.toml', 'requirements/ci/requirements.txt') }}
-          restore-keys: |
-            pip-gpu-
-
       - name: Cache HuggingFace models
         uses: actions/cache@v5
         with:
@@ -65,7 +57,8 @@ jobs:
       - name: Install dependencies
         run: |
           pip install -r requirements/ci/requirements.txt
-          pip install onnxruntime-gpu
+          pip install soundfile librosa ml_dtypes flatbuffers numpy packaging protobuf sympy coloredlogs
+          pip install --pre --index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-13-nightly/pypi/simple/ onnxruntime-gpu onnxruntime-genai-cuda --no-deps
           pip install -e '.[testing,transformers]'
 
       - name: Check for golden test data

.github/workflows/gpu_l5_generation_e2e.yml

@@ -43,14 +43,6 @@ jobs:
         with:
           python-version: "3.12"
 
-      - name: Cache pip packages
-        uses: actions/cache@v5
-        with:
-          path: ~/.cache/pip
-          key: pip-gpu-${{ hashFiles('pyproject.toml', 'requirements/ci/requirements.txt') }}
-          restore-keys: |
-            pip-gpu-
-
       - name: Cache HuggingFace models
         uses: actions/cache@v5
         with:
@@ -65,7 +57,8 @@ jobs:
       - name: Install dependencies
         run: |
           pip install -r requirements/ci/requirements.txt
-          pip install onnxruntime-gpu
+          pip install soundfile librosa ml_dtypes flatbuffers numpy packaging protobuf sympy coloredlogs
+          pip install --pre --index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-13-nightly/pypi/simple/ onnxruntime-gpu onnxruntime-genai-cuda --no-deps
           pip install -e '.[testing,transformers]'
 
       - name: Run L5 generation E2E tests

.github/workflows/main.yml

@@ -270,7 +270,8 @@ jobs:
       - name: Install dependencies
         run: |
           pip install -r requirements/ci/requirements.txt
-          pip install onnxruntime-gpu
+          pip install soundfile librosa ml_dtypes flatbuffers numpy packaging protobuf sympy coloredlogs
+          pip install --pre --index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-13-nightly/pypi/simple/ onnxruntime-gpu onnxruntime-genai-cuda --no-deps
           pip install -e '.[testing]'
       - name: Run fast integration tests
         env:

.github/workflows/validation_examples_gpu.yml

@@ -67,14 +67,6 @@ jobs:
         with:
           python-version: "3.12"
 
-      - name: Cache pip packages
-        uses: actions/cache@v5
-        with:
-          path: ~/.cache/pip
-          key: pip-examples-gpu-${{ hashFiles('pyproject.toml', 'requirements/ci/requirements.txt') }}
-          restore-keys: |
-            pip-examples-gpu-
-
       - name: Cache HuggingFace models
         uses: actions/cache@v5
         with:
@@ -89,9 +81,9 @@ jobs:
       - name: Install dependencies
         run: |
           pip install -r requirements/ci/requirements.txt
-          pip install -e '.[testing,transformers,ort-genai]'
-          pip install onnxruntime-gpu
-          pip install soundfile librosa ml_dtypes
+          pip install -e '.[testing,transformers]'
+          pip install soundfile librosa ml_dtypes flatbuffers numpy packaging protobuf sympy coloredlogs
+          pip install --pre --index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-13-nightly/pypi/simple/ onnxruntime-gpu onnxruntime-genai-cuda --no-deps
 
       - name: Run ${{ matrix.name }}
         env: