Merge branch 'main' into GraniteCodeSupport

Author: Jack-Khuu

.github/workflows/more-tests.yml

@@ -9,23 +9,17 @@ on:
 
 jobs:
   test-cuda:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       timeout: 60
       script: |
         echo "::group::Print machine info"
         uname -a
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
-
         echo "::group::Download checkpoints"
         # Install requirements
         ./install/install_requirements.sh cuda

.github/workflows/periodic.yml

@@ -108,7 +108,7 @@ jobs:
           set -eux
           PYTHONPATH="${PWD}" python .ci/scripts/gather_test_models.py --event "periodic" --backend "gpu"
   test-gpu:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     secrets: inherit
@@ -119,7 +119,7 @@ jobs:
       secrets-env: "HF_TOKEN_PERIODIC"
       runner: ${{ matrix.runner }}
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       script: |
         echo "::group::Print machine info"
         nvidia-smi

.github/workflows/pull.yml

@@ -215,7 +215,7 @@ jobs:
           set -eux
           PYTHONPATH="${PWD}" python .ci/scripts/gather_test_models.py --event "pull_request" --backend "gpu"
   test-gpu-compile:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu-compile (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     strategy:
@@ -224,7 +224,7 @@ jobs:
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       script: |
         echo "::group::Print machine info"
         nvidia-smi
@@ -250,7 +250,7 @@ jobs:
         echo "::endgroup::"
 
   test-gpu-aoti-bfloat16:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu-aoti-bfloat16 (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     strategy:
@@ -259,18 +259,13 @@ jobs:
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       timeout: 60
       script: |
         echo "::group::Print machine info"
         nvidia-smi
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Install required packages"
         ./install/install_requirements.sh cuda
         pip3 list
@@ -291,7 +286,7 @@ jobs:
         echo "::endgroup::"
 
   test-gpu-aoti-float32:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu-aoti-float32 (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     strategy:
@@ -300,17 +295,12 @@ jobs:
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       script: |
         echo "::group::Print machine info"
         nvidia-smi
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Install required packages"
         ./install/install_requirements.sh cuda
         pip list
@@ -337,7 +327,7 @@ jobs:
         echo "::endgroup::"
 
   test-gpu-aoti-float16:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu-aoti-float16 (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     strategy:
@@ -346,17 +336,12 @@ jobs:
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       script: |
         echo "::group::Print machine info"
         nvidia-smi
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Install required packages"
         ./install/install_requirements.sh cuda
         pip list
@@ -384,7 +369,7 @@ jobs:
         echo "::endgroup::"
 
   test-gpu-eval-sanity-check:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     name: test-gpu-eval-sanity-check (${{ matrix.platform }}, ${{ matrix.model_name }})
     needs: gather-models-gpu
     strategy:
@@ -393,17 +378,12 @@ jobs:
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       script: |
         echo "::group::Print machine info"
         nvidia-smi
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Install required packages"
         ./install/install_requirements.sh cuda
         pip3 list
@@ -1031,7 +1011,7 @@ jobs:
           echo "Tests complete."
 
   test-build-runner-et-android:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     with:
       runner: linux.4xlarge
       script: |
@@ -1124,3 +1104,41 @@ jobs:
           echo "Generate AOTI"
           python torchchat.py generate stories110M --aoti-package-path ./model.pt2 --prompt "${PRMT}"
           echo "Tests complete."
+
+  test-torchao-experimental-mps:
+    strategy:
+      matrix:
+        runner: [macos-m1-stable]
+    runs-on: ${{matrix.runner}}
+    steps:
+      - name: Checkout repo
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+      - name: Setup Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: 3.10.11
+      - name: Print machine info
+        run: |
+          uname -a
+          if [ $(uname -s) == Darwin ]; then
+            sysctl machdep.cpu.brand_string
+            sysctl machdep.cpu.core_count
+          fi
+      - name: Install torchchat
+        run: |
+          echo "Intalling pip3 packages"
+          ./install/install_requirements.sh
+          pip3 list
+          python3 -c 'import torch;print(f"torch: {torch.__version__, torch.version.git_version}")'
+      - name: Install torchao-ops-mps
+        id: install-torchao-ops-mps
+        run: |
+          bash torchchat/utils/scripts/build_torchao_ops.sh mps
+      - name: Run inference
+        run: |
+          python torchchat.py download stories110M
+          export PRMT="Once upon a time in a land far away"
+          echo "Generate eager"
+          python torchchat.py generate stories110M --temperature 0 --prompt "${PRMT}" --device mps --dtype float32 --quantize '{"linear:afpwx": {"bitwidth": 3, "groupsize": 32}}'

.github/workflows/run-readme-periodic.yml

@@ -10,24 +10,19 @@ on:
 
 jobs:
   test-readme:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     secrets: inherit
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       secrets-env: "HF_TOKEN_PERIODIC"
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       timeout: 60
       script: |
         echo "::group::Print machine info"
         uname -a
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Create script to run README"
         python3 torchchat/utils/scripts/updown.py --create-sections --file README.md > ./run-readme.sh
         # for good measure, if something happened to updown processor,
@@ -44,23 +39,18 @@ jobs:
 
 
   test-quantization-any:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       secrets: inherit
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       timeout: 60
       script: |
         echo "::group::Print machine info"
         uname -a
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Create script to run quantization"
         python3 torchchat/utils/scripts/updown.py --create-sections --file docs/quantization.md > ./run-quantization.sh
         # for good measure, if something happened to updown processor,
@@ -76,24 +66,19 @@ jobs:
         echo "::endgroup::"
 
   test-gguf-any:
-    uses: pytorch/test-infra/.github/workflows/linux_job.yml@main
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     secrets: inherit
     with:
       runner: linux.g5.4xlarge.nvidia.gpu
       secrets-env: "HF_TOKEN_PERIODIC"
       gpu-arch-type: cuda
-      gpu-arch-version: "12.1"
+      gpu-arch-version: "12.4"
       timeout: 60
       script: |
         echo "::group::Print machine info"
         uname -a
         echo "::endgroup::"
 
-        echo "::group::Install newer objcopy that supports --set-section-alignment"
-        yum install -y  devtoolset-10-binutils
-        export PATH=/opt/rh/devtoolset-10/root/usr/bin/:$PATH
-        echo "::endgroup::"
-
         echo "::group::Create script to run gguf"
         python3 torchchat/utils/scripts/updown.py --file docs/GGUF.md > ./run-gguf.sh
         # for good measure, if something happened to updown processor,