Update on "Use generated files in module test"

Instead of checking in the PTE/PTD files. This helps us keep up to date with aot changes, especially with incoming flat_tensor refactor.

Main changes here are with the model. Instead of using AddModule from examples/portable, use ModuleAdd from test/models/export_program, like our other tests. There is a shape difference (1 tensor vs 2 tensors+alpha), so we have to change all the inputs.

Differential Revision: [D73470865](https://our.internmc.facebook.com/intern/diff/D73470865/)

[ghstack-poisoned]

Author: lucylq

.ci/scripts/gather_benchmark_configs.py

@@ -21,6 +21,7 @@
     "apple_iphone_15": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/3b5acd2e-92e2-4778-b651-7726bafe129d",
     "apple_iphone_15+ios_18": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/12c8b15c-8d03-4e07-950d-0a627e7595b4",
     "samsung_galaxy_s22": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/e59f866a-30aa-4aa1-87b7-4510e5820dfa",
+    "samsung_galaxy_s22_private": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/ea6b049d-1508-4233-9a56-5d9eacbe1078",
     "samsung_galaxy_s24": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/98f8788c-2e25-4a3c-8bb2-0d1e8897c0db",
     "google_pixel_8_pro": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/d65096ab-900b-4521-be8b-a3619b69236a",
     "google_pixel_3_private_rooted": "arn:aws:devicefarm:us-west-2:308535385114:devicepool:02a2cf0f-6d9b-45ee-ba1a-a086587469e6/98d23ca8-ea9e-4fb7-b725-d402017b198d",

.github/scripts/extract_benchmark_results.py

@@ -349,7 +349,10 @@ def transform(
     # Overwrite the device name here with the job name as it has more information about
     # the device, i.e. Samsung Galaxy S22 5G instead of just Samsung
     for r in benchmark_results:
-        r["deviceInfo"]["device"] = job_name
+        is_private_device = job_report.get("is_private_instance", False)
+        r["deviceInfo"]["device"] = (
+            f"{job_name} (private)" if is_private_device else job_name
+        )
 
     # From https://github.com/pytorch/pytorch/wiki/How-to-integrate-with-PyTorch-OSS-benchmark-database
     return [
@@ -363,6 +366,7 @@ def transform(
                     "benchmark_config": json.dumps(benchmark_config),
                     "job_conclusion": "SUCCESS",
                     "job_arn": job_report.get("arn", ""),
+                    "instance_arn": job_report.get("instance_arn", ""),
                 },
             },
             "model": {

.github/workflows/android-perf-private-device-experiment.yml

@@ -23,7 +23,7 @@ on:
         description: Target devices to run benchmark
         required: false
         type: string
-        default: google_pixel_3_private_rooted
+        default: samsung_galaxy_s22_private
       benchmark_configs:
         description: The list of configs used the benchmark
         required: false
@@ -39,7 +39,7 @@ on:
         description: Target devices to run benchmark
         required: false
         type: string
-        default: google_pixel_3_private_rooted
+        default: samsung_galaxy_s22_private
       benchmark_configs:
         description: The list of configs used the benchmark
         required: false
@@ -58,5 +58,5 @@ jobs:
       contents: read
     with:
       models: ${{ inputs.models || 'mv3,meta-llama/Llama-3.2-1B-Instruct-SpinQuant_INT4_EO8,meta-llama/Llama-3.2-1B-Instruct-QLORA_INT4_EO8' }}
-      devices: google_pixel_3_private_rooted
+      devices: samsung_galaxy_s22_private
       benchmark_configs: ${{ inputs.benchmark_configs }}

.github/workflows/apple-perf-private-device-experiment.yml

@@ -1,18 +1,16 @@
 name: apple-perf (private devices)
 
 on:
-  # TODO (huydhn): Disable the schedule run until we land the change to add device pool and device name
-  # to separate between public and private iOS devices
-  # schedule:
-  # - cron: 0 0,4,8,12,16,20 * * *
+  schedule:
+   - cron: 0 0,4,8,12,16,20 * * *
   pull_request:
     paths:
       - .github/workflows/apple-perf-private-device-experiment.yml
-  # push:
-  #   branches:
-  #     - main
-  #   paths:
-  #     - .github/workflows/apple-perf-private-device-experiment.yml
+  push:
+    branches:
+      - main
+    paths:
+      - .github/workflows/apple-perf-private-device-experiment.yml
   # Note: GitHub has an upper limit of 10 inputs
   workflow_dispatch:
     inputs:

.github/workflows/doc-build.yml

@@ -14,20 +14,6 @@ on:
     - cron: '0 0 * * *'
 
 jobs:
-  check-urls:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v3
-      - name: Check URLs
-        run: bash ./scripts/check_urls.sh
-
-  check-xrefs:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v3
-      - name: Check Links
-        run: bash ./scripts/check_xrefs.sh
-
   build:
     uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     permissions:

.github/workflows/lint.yml

@@ -64,6 +64,30 @@ jobs:
 
         exit $RC
 
+  lint-urls:
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
+    with:
+      runner: linux.2xlarge
+      docker-image: executorch-ubuntu-22.04-linter
+      submodules: 'none'
+      fetch-depth: 0
+      ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
+      timeout: 90
+      script: |
+        ./scripts/lint_urls.sh
+
+  lint-xrefs:
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
+    with:
+      runner: linux.2xlarge
+      docker-image: executorch-ubuntu-22.04-linter
+      submodules: 'none'
+      fetch-depth: 0
+      ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
+      timeout: 90
+      script: |
+        ./scripts/lint_xrefs.sh
+
   android-java-format:
     uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     permissions:

.github/workflows/pull.yml

@@ -481,37 +481,38 @@ jobs:
       build-tool: buck2
       docker-image: executorch-ubuntu-22.04-clang12
 
-  unittest-arm:
+  unittest-arm-backend-with-no-fvp:
+    name: unittest-arm-backend-with-no-fvp
     uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     permissions:
       id-token: write
       contents: read
+    strategy:
+      matrix:
+        include:
+          - test_arm_baremetal: test_pytest_ops
+          - test_arm_baremetal: test_pytest_models
+      fail-fast: false
     with:
       runner: linux.2xlarge
       docker-image: executorch-ubuntu-22.04-arm-sdk
       submodules: 'recursive'
       ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
       timeout: 90
       script: |
-        set -eux
-
         # The generic Linux job chooses to use base env, not the one setup by the image
         CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
         conda activate "${CONDA_ENV}"
 
-        BUILD_TOOL="cmake"
-
-        # Setup MacOS dependencies as there is no Docker support on MacOS atm
-        PYTHON_EXECUTABLE=python \
-        CMAKE_ARGS="-DEXECUTORCH_BUILD_PYBIND=ON" \
-        EXECUTORCH_BUILD_ARM_BAREMETAL=ON \
-        .ci/scripts/setup-linux.sh --build-tool "${BUILD_TOOL}"
+        source .ci/scripts/utils.sh
+        install_executorch "--use-pt-pinned-commit"
 
-        # Install Arm dependencies
         .ci/scripts/setup-arm-baremetal-tools.sh
 
-        # Run pytest without simulator
-        backends/arm/test/test_arm_baremetal.sh test_pytest
+        ARM_TEST=${{ matrix.test_arm_baremetal }}
+
+        # Test test_arm_baremetal.sh with test
+        backends/arm/test/test_arm_baremetal.sh "${ARM_TEST}"
 
   test-llama-runner-qnn-linux:
     name: test-llama-runner-qnn-linux

.github/workflows/trunk.yml

@@ -176,12 +176,22 @@ jobs:
         # Test selective build
         PYTHON_EXECUTABLE=python bash examples/portable/scripts/test_demo_backend_delegation.sh "${BUILD_TOOL}"
 
-  test-arm-backend-delegation:
-    name: test-arm-backend-delegation
+  test-arm-backend:
+    name: test-arm-backend
     uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
     permissions:
       id-token: write
       contents: read
+    strategy:
+      matrix:
+        include:
+          - test_arm_baremetal: test_pytest_ops_ethosu_fvp
+          - test_arm_baremetal: test_pytest_models_ethosu_fvp
+          - test_arm_baremetal: test_run_ethosu_fvp
+          - test_arm_baremetal: test_models_tosa
+          - test_arm_baremetal: test_models_ethos-u55
+          - test_arm_baremetal: test_models_ethos-u85
+      fail-fast: false
     with:
       runner: linux.2xlarge.memory
       docker-image: executorch-ubuntu-22.04-arm-sdk
@@ -202,34 +212,10 @@ jobs:
         # Hopefully this is high enough for this setup.
         sudo sysctl fs.inotify.max_user_watches=1048576 # 1024 * 1024
 
-        # Test ethos-u delegate examples with run.sh
-        backends/arm/test/test_arm_baremetal.sh test_full_ethosu_fvp
-
-
-  test-arm-reference-delegation:
-    name: test-arm-reference-delegation
-    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
-    permissions:
-      id-token: write
-      contents: read
-    with:
-      runner: linux.2xlarge.memory
-      docker-image: executorch-ubuntu-22.04-arm-sdk
-      submodules: 'recursive'
-      ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
-      timeout: 90
-      script: |
-        # The generic Linux job chooses to use base env, not the one setup by the image
-        CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
-        conda activate "${CONDA_ENV}"
-
-        source .ci/scripts/utils.sh
-        install_executorch "--use-pt-pinned-commit"
-
-        .ci/scripts/setup-arm-baremetal-tools.sh
+        ARM_TEST=${{ matrix.test_arm_baremetal }}
 
-        # Run arm unit tests using the simulator
-        backends/arm/test/test_arm_baremetal.sh test_pytest_ethosu_fvp
+        # Test test_arm_baremetal.sh with test
+        backends/arm/test/test_arm_baremetal.sh "${ARM_TEST}"
 
   test-arm-cortex-m-size-test:
     name: test-arm-cortex-m-size-test

.lintrunner.toml

@@ -10,7 +10,7 @@ exclude_patterns = [
     'exir/serde/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -19,7 +19,7 @@ command = [
     '@{{PATHSFILE}}'
 ]
 init_command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -41,7 +41,7 @@ exclude_patterns = [
     'exir/serde/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -50,7 +50,7 @@ command = [
     '@{{PATHSFILE}}'
 ]
 init_command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -83,7 +83,7 @@ exclude_patterns = [
     'runtime/core/portable_type/c10/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -94,7 +94,7 @@ command = [
     '@{{PATHSFILE}}'
 ]
 init_command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -116,7 +116,7 @@ exclude_patterns = [
     '**/third-party/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -126,7 +126,7 @@ command = [
     '@{{PATHSFILE}}',
 ]
 init_command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -150,7 +150,7 @@ exclude_patterns = [
     '**/third-party/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -191,7 +191,7 @@ exclude_patterns = [
     'extension/llm/custom_ops/spinquant/test/fast_hadamard_transform_special_unstrided_cpu.h',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -226,7 +226,7 @@ exclude_patterns = [
     'util/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -275,7 +275,7 @@ exclude_patterns = [
     'util/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -325,7 +325,7 @@ exclude_patterns = [
     'backends/arm/test/**',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -337,7 +337,7 @@ command = [
     '@{{PATHSFILE}}'
 ]
 init_command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',
@@ -356,7 +356,7 @@ exclude_patterns = [
     '.lintrunner.toml',
 ]
 command = [
-    'python',
+    'python3',
     '-m',
     'lintrunner_adapters',
     'run',

README.md

@@ -49,8 +49,8 @@ Key value propositions of ExecuTorch are:
 ## Getting Started
 To get started you can:
 
-- Visit the [Step by Step Tutorial](https://pytorch.org/executorch/main/index) to get things running locally and deploy a model to a device
-- Use this [Colab Notebook](https://pytorch.org/executorch/main/getting-started-setup#quick-setup-colab-jupyter-notebook-prototype) to start playing around right away
+- Visit the [Step by Step Tutorial](https://pytorch.org/executorch/stable/getting-started.html) to get things running locally and deploy a model to a device
+- Use this [Colab Notebook](https://colab.research.google.com/drive/1qpxrXC3YdJQzly3mRg-4ayYiOjC6rue3?usp=sharing) to start playing around right away
 - Jump straight into LLM use cases by following specific instructions for [Llama](examples/models/llama/README.md) and [Llava](examples/models/llava/README.md)
 
 ## Feedback and Engagement