Fix deserialization of TransformerEncoderLayer (#81832) (#81832) (#82094)

Summary:
When `activation` is a module, it is not saved directly in the state dictionary but instead in `_modules`. When deserialized, the old version of this code would think that activation was missing and set it to RELU. This version first reconstructions the module and then sees if activation is neither a module nor a function before setting it to RELU.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/81832
Approved by: https://github.com/kit1980, https://github.com/zrphercule

Test Plan:
contbuild & OSS CI, see e68583b4d1

Test plan from GitHub:
pytorch oss tests

Reviewed By: jeanschmidt, zrphercule

Differential Revision: D38014872

Pulled By: zdevito

fbshipit-source-id: 938079d768f7981ca55eed3c8828b29a92e06f41

Co-authored-by: Zachary DeVito (Meta Employee) <zdevito@fb.com>

.bazelrc

@@ -13,103 +13,15 @@ build:no-tty --curses no
 build:no-tty --progress_report_interval 10
 build:no-tty --show_progress_rate_limit 10
 
-# Build with GPU support by default.
-build --define=cuda=true
-# rules_cuda configuration
-build --@rules_cuda//cuda:enable_cuda
-build --@rules_cuda//cuda:cuda_targets=sm_52
-build --@rules_cuda//cuda:compiler=nvcc
-build --repo_env=CUDA_PATH=/usr/local/cuda
-
-# Configuration to build without GPU support
-build:cpu-only --define=cuda=false
+# Configuration to build with GPU support
+build:gpu --define=cuda=true
 # define a separate build folder for faster switching between configs
-build:cpu-only --platform_suffix=-cpu-only
+build:gpu --platform_suffix=-gpu
 # See the note on the config-less build for details about why we are
-# doing this. We must also do it for the "-cpu-only" platform suffix.
-build --copt=-isystem --copt=bazel-out/k8-fastbuild-cpu-only/bin
+# doing this. We must also do it for the "-gpu" platform suffix.
+build --copt=-isystem --copt=bazel-out/k8-fastbuild-gpu/bin
 # rules_cuda configuration
-build:cpu-only --@rules_cuda//cuda:enable_cuda=False
-
-# Definition of --config=shell
-# interactive shell immediately before execution
-build:shell --run_under="//tools/bazel_tools:shellwrap"
-
-# Disable all warnings for external repositories. We don't care about
-# their warnings.
-build --per_file_copt=^external/@-w
-
-# Set additional warnings to error level.
-#
-# Implementation notes:
-#  * we use file extensions to determine if we are using the C++
-#    compiler or the cuda compiler
-#  * we use ^// at the start of the regex to only permit matching
-#    PyTorch files. This excludes external repos.
-#
-# Note that because this is logically a command-line flag, it is
-# considered the word on what warnings are enabled. This has the
-# unfortunate consequence of preventing us from disabling an error at
-# the target level because those flags will come before these flags in
-# the action invocation. Instead we provide per-file exceptions after
-# this.
-#
-# On the bright side, this means we don't have to more broadly apply
-# the exceptions to an entire target.
-#
-# Looking for CUDA flags? We have a cu_library macro that we can edit
-# directly. Look in //tools/rules:cu.bzl for details. Editing the
-# macro over this has the following advantages:
-#  * making changes does not require discarding the Bazel analysis
-#    cache
-#  * it allows for selective overrides on individual targets since the
-#    macro-level opts will come earlier than target level overrides
-
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Werror=all
-# The following warnings come from -Wall. We downgrade them from error
-# to warnings here.
-#
-# sign-compare has a tremendous amount of violations in the
-# codebase. It will be a lot of work to fix them, just disable it for
-# now.
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Wno-sign-compare
-# We intentionally use #pragma unroll, which is compiler specific.
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Wno-error=unknown-pragmas
-
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Werror=extra
-# The following warnings come from -Wextra. We downgrade them from error
-# to warnings here.
-#
-# unused-parameter-compare has a tremendous amount of violations in the
-# codebase. It will be a lot of work to fix them, just disable it for
-# now.
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Wno-unused-parameter
-# missing-field-parameters has both a large number of violations in
-# the codebase, but it also is used pervasively in the Python C
-# API. There are a couple of catches though:
-# * we use multiple versions of the Python API and hence have
-#   potentially multiple different versions of each relevant
-#   struct. They may have different numbers of fields. It will be
-#   unwieldy to support multiple versions in the same source file.
-# * Python itself for many of these structs recommends only
-#   initializing a subset of the fields. We should respect the API
-#   usage conventions of our dependencies.
-#
-# Hence, we just disable this warning altogether. We may want to clean
-# up some of the clear-cut cases that could be risky, but we still
-# likely want to have this disabled for the most part.
-build --per_file_copt='^//.*\.(cpp|cc)$'@-Wno-missing-field-initializers
-
-build --per_file_copt='//:aten/src/ATen/RegisterCompositeExplicitAutograd\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterCompositeImplicitAutograd\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterMkldnnCPU\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterNestedTensorCPU\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterQuantizedCPU\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterSparseCPU\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterSparseCsrCPU\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterNestedTensorMeta\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterSparseMeta\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterQuantizedMeta\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:aten/src/ATen/RegisterZeroTensor\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:torch/csrc/lazy/generated/RegisterAutogradLazy\.cpp$'@-Wno-error=unused-function
-build --per_file_copt='//:torch/csrc/lazy/generated/RegisterLazy\.cpp$'@-Wno-error=unused-function
+build:gpu --@rules_cuda//cuda:enable_cuda
+build:gpu --@rules_cuda//cuda:cuda_targets=sm_52
+build:gpu --@rules_cuda//cuda:compiler=nvcc
+build:gpu --repo_env=CUDA_PATH=/usr/local/cuda

.buckconfig.oss

@@ -1,13 +1,8 @@
-[pt]
-  is_oss=1
-
 [buildfile]
-  name = BUCK.oss
-  includes = //tools/build_defs/select.bzl
+name = BUILD.buck
 
 [repositories]
   bazel_skylib = third_party/bazel-skylib/
-  ovr_config = .
 
 [download]
   in_build = true
@@ -15,11 +10,6 @@
 [cxx]
   cxxflags = -std=c++17
   should_remap_host_platform = true
-  cpp = /usr/bin/clang
-  cc = /usr/bin/clang
-  cxx = /usr/bin/clang++
-  cxxpp = /usr/bin/clang++
-  ld = /usr/bin/clang++
 
 [project]
   default_flavors_mode=all

.circleci/cimodel/data/dimensions.py

@@ -4,7 +4,6 @@ CUDA_VERSIONS = [
     "102",
     "113",
     "116",
-    "117",
 ]
 
 ROCM_VERSIONS = [

.circleci/cimodel/data/pytorch_build_data.py

@@ -75,7 +75,6 @@ class ExperimentalFeatureConfigNode(TreeConfigNode):
             "vulkan": VulkanConfigNode,
             "parallel_tbb": ParallelTBBConfigNode,
             "crossref": CrossRefConfigNode,
-            "dynamo": DynamoConfigNode,
             "parallel_native": ParallelNativeConfigNode,
             "onnx": ONNXConfigNode,
             "libtorch": LibTorchConfigNode,
@@ -180,14 +179,6 @@ class CrossRefConfigNode(TreeConfigNode):
         return ImportantConfigNode
 
 
-class DynamoConfigNode(TreeConfigNode):
-    def init2(self, node_name):
-        self.props["is_dynamo"] = node_name
-
-    def child_constructor(self):
-        return ImportantConfigNode
-
-
 class ParallelNativeConfigNode(TreeConfigNode):
     def modify_label(self, label):
         return "PARALLELNATIVE=" + str(label)

.circleci/cimodel/data/pytorch_build_definitions.py

@@ -240,7 +240,6 @@ def instantiate_configs(only_slow_gradcheck):
         is_xla = fc.find_prop("is_xla") or False
         is_asan = fc.find_prop("is_asan") or False
         is_crossref = fc.find_prop("is_crossref") or False
-        is_dynamo = fc.find_prop("is_dynamo") or False
         is_onnx = fc.find_prop("is_onnx") or False
         is_pure_torch = fc.find_prop("is_pure_torch") or False
         is_vulkan = fc.find_prop("is_vulkan") or False
@@ -287,9 +286,6 @@ def instantiate_configs(only_slow_gradcheck):
         if is_crossref:
             parms_list_ignored_for_docker_image.append("crossref")
 
-        if is_dynamo:
-            parms_list_ignored_for_docker_image.append("dynamo")
-
         if is_onnx:
             parms_list.append("onnx")
             python_version = fc.find_prop("pyver")

.circleci/cimodel/data/simple/ios_definitions.py

@@ -1,5 +1,4 @@
 from cimodel.data.simple.util.versions import MultiPartVersion
-from cimodel.data.simple.util.branch_filters import gen_filter_dict_exclude
 import cimodel.lib.miniutils as miniutils
 
 XCODE_VERSION = MultiPartVersion([12, 5, 1])
@@ -12,7 +11,7 @@ class ArchVariant:
 
     def render(self):
         extra_parts = [self.custom_build_name] if len(self.custom_build_name) > 0 else []
-        return "-".join([self.name] + extra_parts).replace("_", "-")
+        return "_".join([self.name] + extra_parts)
 
 
 def get_platform(arch_variant_name):
@@ -26,25 +25,30 @@ class IOSJob:
         self.is_org_member_context = is_org_member_context
         self.extra_props = extra_props
 
-    def gen_name_parts(self):
-        version_parts = self.xcode_version.render_dots_or_parts("-")
-        build_variant_suffix = self.arch_variant.render()
+    def gen_name_parts(self, with_version_dots):
+
+        version_parts = self.xcode_version.render_dots_or_parts(with_version_dots)
+        build_variant_suffix = "_".join([self.arch_variant.render(), "build"])
+
         return [
+            "pytorch",
             "ios",
         ] + version_parts + [
             build_variant_suffix,
         ]
 
     def gen_job_name(self):
-        return "-".join(self.gen_name_parts())
+        return "_".join(self.gen_name_parts(False))
 
     def gen_tree(self):
+
         platform_name = get_platform(self.arch_variant.name)
+
         props_dict = {
-            "name": self.gen_job_name(),
-            "build_environment": self.gen_job_name(),
+            "build_environment": "-".join(self.gen_name_parts(True)),
             "ios_arch": self.arch_variant.name,
             "ios_platform": platform_name,
+            "name": self.gen_job_name(),
         }
 
         if self.is_org_member_context:
@@ -53,28 +57,30 @@ class IOSJob:
         if self.extra_props:
             props_dict.update(self.extra_props)
 
-        props_dict["filters"] = gen_filter_dict_exclude()
-
         return [{"pytorch_ios_build": props_dict}]
 
 
 WORKFLOW_DATA = [
     IOSJob(XCODE_VERSION, ArchVariant("x86_64"), is_org_member_context=False, extra_props={
         "lite_interpreter": miniutils.quote(str(int(True)))}),
-    # IOSJob(XCODE_VERSION, ArchVariant("arm64"), extra_props={
-    #     "lite_interpreter": miniutils.quote(str(int(True)))}),
-    # IOSJob(XCODE_VERSION, ArchVariant("arm64", "metal"), extra_props={
-    #     "use_metal": miniutils.quote(str(int(True))),
-    #     "lite_interpreter": miniutils.quote(str(int(True)))}),
-    # IOSJob(XCODE_VERSION, ArchVariant("arm64", "custom-ops"), extra_props={
-    #     "op_list": "mobilenetv2.yaml",
-    #     "lite_interpreter": miniutils.quote(str(int(True)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("x86_64", "full_jit"), is_org_member_context=False, extra_props={
+        "lite_interpreter": miniutils.quote(str(int(False)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("arm64"), extra_props={
+        "lite_interpreter": miniutils.quote(str(int(True)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("arm64", "metal"), extra_props={
+        "use_metal": miniutils.quote(str(int(True))),
+        "lite_interpreter": miniutils.quote(str(int(True)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("arm64", "full_jit"), extra_props={
+        "lite_interpreter": miniutils.quote(str(int(False)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("arm64", "custom"), extra_props={
+        "op_list": "mobilenetv2.yaml",
+        "lite_interpreter": miniutils.quote(str(int(True)))}),
     IOSJob(XCODE_VERSION, ArchVariant("x86_64", "coreml"), is_org_member_context=False, extra_props={
         "use_coreml": miniutils.quote(str(int(True))),
         "lite_interpreter": miniutils.quote(str(int(True)))}),
-    # IOSJob(XCODE_VERSION, ArchVariant("arm64", "coreml"), extra_props={
-    #     "use_coreml": miniutils.quote(str(int(True))),
-    #     "lite_interpreter": miniutils.quote(str(int(True)))}),
+    IOSJob(XCODE_VERSION, ArchVariant("arm64", "coreml"), extra_props={
+        "use_coreml": miniutils.quote(str(int(True))),
+        "lite_interpreter": miniutils.quote(str(int(True)))}),
 ]
 
 

.circleci/cimodel/data/simple/macos_definitions.py

@@ -1,8 +1,3 @@
-from collections import OrderedDict
-from cimodel.lib.miniutils import quote
-from cimodel.data.simple.util.branch_filters import gen_filter_dict_exclude
-
-
 class MacOsJob:
     def __init__(self, os_version, is_build=False, is_test=False, extra_props=tuple()):
         # extra_props is tuple type, because mutable data structures for argument defaults
@@ -16,14 +11,10 @@ class MacOsJob:
         non_phase_parts = ["pytorch", "macos", self.os_version, "py3"]
 
         extra_name_list = [name for name, exist in self.extra_props.items() if exist]
-        full_job_name_list = (
-            non_phase_parts
-            + extra_name_list
-            + [
-                "build" if self.is_build else None,
-                "test" if self.is_test else None,
-            ]
-        )
+        full_job_name_list = non_phase_parts + extra_name_list + [
+            'build' if self.is_build else None,
+            'test' if self.is_test else None,
+        ]
 
         full_job_name = "_".join(list(filter(None, full_job_name_list)))
 
@@ -50,99 +41,12 @@ WORKFLOW_DATA = [
         "10_13",
         is_build=True,
         is_test=True,
-        extra_props=tuple({"lite_interpreter": True}.items()),
-    ),
+        extra_props=tuple({
+            "lite_interpreter": True
+        }.items()),
+    )
 ]
 
 
-def get_new_workflow_jobs():
-    return [
-        OrderedDict(
-            {
-                "mac_build": OrderedDict(
-                    {
-                        "name": "macos-12-py3-x86-64-build",
-                        "build-environment": "macos-12-py3-x86-64",
-                        "xcode-version": quote("13.3.1"),
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-        OrderedDict(
-            {
-                "mac_test": OrderedDict(
-                    {
-                        "name": "macos-12-py3-x86-64-test-1-2-default",
-                        "build-environment": "macos-12-py3-x86-64",
-                        "xcode-version": quote("13.3.1"),
-                        "shard-number": quote("1"),
-                        "num-test-shards": quote("2"),
-                        "requires": ["macos-12-py3-x86-64-build"],
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-        OrderedDict(
-            {
-                "mac_test": OrderedDict(
-                    {
-                        "name": "macos-12-py3-x86-64-test-2-2-default",
-                        "build-environment": "macos-12-py3-x86-64",
-                        "xcode-version": quote("13.3.1"),
-                        "shard-number": quote("2"),
-                        "num-test-shards": quote("2"),
-                        "requires": ["macos-12-py3-x86-64-build"],
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-        OrderedDict(
-            {
-                "mac_test": OrderedDict(
-                    {
-                        "name": "macos-12-py3-x86-64-test-1-1-functorch",
-                        "build-environment": "macos-12-py3-x86-64",
-                        "xcode-version": quote("13.3.1"),
-                        "shard-number": quote("1"),
-                        "num-test-shards": quote("1"),
-                        "test-config": "functorch",
-                        "requires": ["macos-12-py3-x86-64-build"],
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-        OrderedDict(
-            {
-                "mac_build": OrderedDict(
-                    {
-                        "name": "macos-12-py3-x86-64-lite-interpreter-build-test",
-                        "build-environment": "macos-12-py3-lite-interpreter-x86-64",
-                        "xcode-version": quote("13.3.1"),
-                        "build-generates-artifacts": "false",
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-        OrderedDict(
-            {
-                "mac_build": OrderedDict(
-                    {
-                        "name": "macos-12-py3-arm64-build",
-                        "build-environment": "macos-12-py3-arm64",
-                        "xcode-version": quote("13.3.1"),
-                        "python-version": quote("3.9.12"),
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-    ]
-
-
 def get_workflow_jobs():
     return [item.gen_tree() for item in WORKFLOW_DATA]

.circleci/cimodel/data/simple/nightly_ios.py

@@ -15,7 +15,7 @@ class IOSNightlyJob:
     def get_phase_name(self):
         return "upload" if self.is_upload else "build"
 
-    def get_common_name_pieces(self, sep):
+    def get_common_name_pieces(self, with_version_dots):
 
         extra_name_suffix = [self.get_phase_name()] if self.is_upload else []
 
@@ -24,7 +24,7 @@ class IOSNightlyJob:
         common_name_pieces = [
             "ios",
         ] + extra_name + [
-        ] + ios_definitions.XCODE_VERSION.render_dots_or_parts(sep) + [
+        ] + ios_definitions.XCODE_VERSION.render_dots_or_parts(with_version_dots) + [
             "nightly",
             self.variant,
             "build",
@@ -33,14 +33,14 @@ class IOSNightlyJob:
         return common_name_pieces
 
     def gen_job_name(self):
-        return "_".join(["pytorch"] + self.get_common_name_pieces(None))
+        return "_".join(["pytorch"] + self.get_common_name_pieces(False))
 
     def gen_tree(self):
         build_configs = BUILD_CONFIGS_FULL_JIT if self.is_full_jit else BUILD_CONFIGS
         extra_requires = [x.gen_job_name() for x in build_configs] if self.is_upload else []
 
         props_dict = {
-            "build_environment": "-".join(["libtorch"] + self.get_common_name_pieces(".")),
+            "build_environment": "-".join(["libtorch"] + self.get_common_name_pieces(True)),
             "requires": extra_requires,
             "context": "org-member",
             "filters": {"branches": {"only": "nightly"}},

.circleci/cimodel/data/simple/upload_test_stats_definition.py

@@ -1,22 +0,0 @@
-from typing import OrderedDict
-from cimodel.data.simple.util.branch_filters import gen_filter_dict_exclude
-
-
-def get_workflow_job():
-    return [
-        OrderedDict(
-            {
-                "upload_test_stats": OrderedDict(
-                    {
-                        "name": "upload test status",
-                        "requires": [
-                            "macos-12-py3-x86-64-test-1-2-default",
-                            "macos-12-py3-x86-64-test-2-2-default",
-                            "macos-12-py3-x86-64-test-1-1-functorch",
-                        ],
-                        "filters": gen_filter_dict_exclude()
-                    }
-                )
-            }
-        ),
-    ]

.circleci/cimodel/data/simple/util/branch_filters.py

@@ -12,9 +12,6 @@ PR_BRANCH_LIST = [
 
 RC_PATTERN = r"/v[0-9]+(\.[0-9]+)*-rc[0-9]+/"
 
-MAC_IOS_EXCLUSION_LIST = ["nightly", "postnightly"]
-
-
 def gen_filter_dict(
         branches_list=NON_PR_BRANCH_LIST,
         tags_list=None
@@ -29,11 +26,3 @@ def gen_filter_dict(
     if tags_list is not None:
         filter_dict["tags"] = {"only": tags_list}
     return filter_dict
-
-
-def gen_filter_dict_exclude(branches_list=MAC_IOS_EXCLUSION_LIST):
-    return {
-        "branches": {
-            "ignore": branches_list,
-        },
-    }

.circleci/cimodel/data/simple/util/versions.py

@@ -1,6 +1,3 @@
-from typing import Optional
-
-
 class MultiPartVersion:
     def __init__(self, parts, prefix=""):
         self.parts = parts
@@ -16,11 +13,14 @@ class MultiPartVersion:
         else:
             return [self.prefix]
 
-    def render_dots_or_parts(self, sep: Optional[str] = None):
-        if sep is None:
-            return self.prefixed_parts()
+    def render_dots(self):
+        return ".".join(self.prefixed_parts())
+
+    def render_dots_or_parts(self, with_dots):
+        if with_dots:
+            return [self.render_dots()]
         else:
-            return [sep.join(self.prefixed_parts())]
+            return self.prefixed_parts()
 
 
 class CudaVersion(MultiPartVersion):

.circleci/config.yml

@@ -174,6 +174,46 @@ commands:
               echo "This is not a pull request, skipping..."
             fi
 
+  upload_binary_size_for_android_build:
+    description: "Upload binary size data for Android build"
+    parameters:
+      build_type:
+        type: string
+        default: ""
+      artifacts:
+        type: string
+        default: ""
+    steps:
+      - run:
+          name: "Binary Size - Install Dependencies"
+          no_output_timeout: "5m"
+          command: |
+            retry () {
+              $* || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
+            }
+            retry pip3 install requests
+      - run:
+          name: "Binary Size - Untar Artifacts"
+          no_output_timeout: "5m"
+          command: |
+            # The artifact file is created inside docker container, which contains the result binaries.
+            # Now unpackage it into the project folder. The subsequent script will scan project folder
+            # to locate result binaries and report their sizes.
+            # If artifact file is not provided it assumes that the project folder has been mounted in
+            # the docker during build and already contains the result binaries, so this step can be skipped.
+            export ARTIFACTS="<< parameters.artifacts >>"
+            if [ -n "${ARTIFACTS}" ]; then
+              tar xf "${ARTIFACTS}" -C ~/project
+            fi
+      - run:
+          name: "Binary Size - Upload << parameters.build_type >>"
+          no_output_timeout: "5m"
+          command: |
+            cd ~/project
+            export ANDROID_BUILD_TYPE="<< parameters.build_type >>"
+            export COMMIT_TIME=$(git log --max-count=1 --format=%ct || echo 0)
+            python3 -m tools.stats.upload_binary_size_to_scuba android
+
 ##############################################################################
 # Binary build (nightlies nightly build) defaults
 # The binary builds use the docker executor b/c at time of writing the machine
@@ -401,6 +441,245 @@ binary_windows_params: &binary_windows_params
 # Job specs
 ##############################################################################
 jobs:
+  binary_linux_build:
+    <<: *binary_linux_build_params
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Build
+        no_output_timeout: "1h"
+        command: |
+            source "/pytorch/.circleci/scripts/binary_linux_build.sh"
+            # Preserve build log
+            if [ -f /pytorch/build/.ninja_log ]; then
+              cp /pytorch/build/.ninja_log /final_pkgs
+            fi
+    - run:
+        name: Output binary sizes
+        no_output_timeout: "1m"
+        command: |
+            ls -lah /final_pkgs
+    - run:
+        name: upload build & binary data
+        no_output_timeout: "5m"
+        command: |
+            source /env
+            cd /pytorch && export COMMIT_TIME=$(git log --max-count=1 --format=%ct || echo 0)
+            python3 -mpip install requests && \
+            SCRIBE_GRAPHQL_ACCESS_TOKEN=${SCRIBE_GRAPHQL_ACCESS_TOKEN} \
+            python3 -m tools.stats.upload_binary_size_to_scuba || exit 0
+    - persist_to_workspace:
+        root: /
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /final_pkgs
+
+    # This should really just be another step of the binary_linux_build job above.
+    # This isn't possible right now b/c the build job uses the docker executor
+    # (otherwise they'd be really really slow) but this one uses the macine
+    # executor (b/c we have to run the docker with --runtime=nvidia and we can't do
+    # that on the docker executor)
+  binary_linux_test:
+    <<: *binary_linux_test_upload_params
+    machine:
+        image: ubuntu-2004:202104-01
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - attach_workspace:
+        at: /home/circleci/project
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Prepare test code
+        no_output_timeout: "1h"
+        command: .circleci/scripts/binary_linux_test.sh
+    - run:
+        <<: *binary_run_in_docker
+
+  binary_upload:
+    parameters:
+      package_type:
+        type: string
+        description: "What type of package we are uploading (eg. wheel, libtorch, conda)"
+        default: "wheel"
+      upload_subfolder:
+        type: string
+        description: "What subfolder to put our package into (eg. cpu, cudaX.Y, etc.)"
+        default: "cpu"
+    docker:
+      - image: continuumio/miniconda3
+    environment:
+      - DRY_RUN: disabled
+      - PACKAGE_TYPE: "<< parameters.package_type >>"
+      - UPLOAD_SUBFOLDER: "<< parameters.upload_subfolder >>"
+    steps:
+      - attach_workspace:
+          at: /tmp/workspace
+      - checkout
+      - designate_upload_channel
+      - run:
+          name: Install dependencies
+          no_output_timeout: "1h"
+          command: |
+            conda install -yq anaconda-client
+            pip install -q awscli
+      - run:
+          name: Do upload
+          no_output_timeout: "1h"
+          command: |
+            AWS_ACCESS_KEY_ID="${PYTORCH_BINARY_AWS_ACCESS_KEY_ID}" \
+              AWS_SECRET_ACCESS_KEY="${PYTORCH_BINARY_AWS_SECRET_ACCESS_KEY}" \
+              ANACONDA_API_TOKEN="${CONDA_PYTORCHBOT_TOKEN}" \
+              .circleci/scripts/binary_upload.sh
+
+  # Nighlty build smoke tests defaults
+  # These are the second-round smoke tests. These make sure that the binaries are
+  # correct from a user perspective, testing that they exist from the cloud are
+  # are runnable. Note that the pytorch repo is never cloned into these jobs
+  ##############################################################################
+  smoke_linux_test:
+    <<: *binary_linux_test_upload_params
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          cat >/home/circleci/project/ci_test_script.sh \<<EOL
+          # The following code will be executed inside Docker container
+          set -eux -o pipefail
+          /builder/smoke_test.sh
+          # The above code will be executed inside Docker container
+          EOL
+    - run:
+        <<: *binary_run_in_docker
+
+  smoke_mac_test:
+    <<: *binary_linux_test_upload_params
+    macos:
+      xcode: "12.0"
+    steps:
+      - checkout
+      - run:
+          <<: *binary_checkout
+      - run:
+          <<: *binary_populate_env
+      - brew_update
+      - run:
+          <<: *binary_install_miniconda
+      - run:
+          name: Build
+          no_output_timeout: "1h"
+          command: |
+            set -ex
+            source "/Users/distiller/project/env"
+            export "PATH=$workdir/miniconda/bin:$PATH"
+            # TODO unbuffer and ts this, but it breaks cause miniconda overwrites
+            # tclsh. But unbuffer and ts aren't that important so they're just
+            # disabled for now
+            ./builder/smoke_test.sh
+
+  binary_mac_build:
+    <<: *binary_mac_params
+    macos:
+      xcode: "12.0"
+      resource_class: "large"
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - brew_update
+    - run:
+        <<: *binary_install_miniconda
+
+    - run:
+        name: Build
+        no_output_timeout: "90m"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_build.sh"
+          cat "$script"
+          source "$script"
+
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_test.sh"
+          cat "$script"
+          source "$script"
+
+    - persist_to_workspace:
+        root: /Users/distiller/project
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /Users/distiller/project/final_pkgs
+
+  binary_macos_arm64_build:
+    <<: *binary_mac_params
+    macos:
+      xcode: "12.3.0"
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - brew_update
+    - run:
+        <<: *binary_install_miniconda
+
+    - run:
+        name: Build
+        no_output_timeout: "90m"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          export CROSS_COMPILE_ARM64=1
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_build.sh"
+          cat "$script"
+          source "$script"
+
+    - persist_to_workspace:
+        root: /Users/distiller/project
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /Users/distiller/project/final_pkgs
+
+
   binary_ios_build:
     <<: *pytorch_ios_params
     macos:
@@ -445,6 +724,90 @@ jobs:
           cat "$script"
           source "$script"
 
+  binary_windows_build:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-xlarge-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Build
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          script="/c/w/p/.circleci/scripts/binary_windows_build.sh"
+          cat "$script"
+          source "$script"
+    - persist_to_workspace:
+        root: "C:/w"
+        paths: final_pkgs
+    - store_artifacts:
+        path: C:/w/final_pkgs
+
+  binary_windows_test:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-medium-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    - checkout
+    - attach_workspace:
+        at: c:/users/circleci/project
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          script="/c/w/p/.circleci/scripts/binary_windows_test.sh"
+          cat "$script"
+          source "$script"
+
+  smoke_windows_test:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-medium-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          export TEST_NIGHTLY_PACKAGE=1
+          script="/c/w/p/.circleci/scripts/binary_windows_test.sh"
+          cat "$script"
+          source "$script"
+
   anaconda_prune:
     parameters:
       packages:
@@ -499,6 +862,95 @@ jobs:
           pushd /tmp/workspace
           git push -u origin "<< parameters.branch >>"
 
+  pytorch_python_doc_build:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-python-doc-push
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: large
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc Build and Push
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          # turn v1.12.0rc3 into 1.12
+          tag=$(echo $CIRCLE_TAG | sed -e 's/v*\([0-9]*\.[0-9]*\).*/\1/')
+          target=${tag:-main}
+          echo "building for ${target}"
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --env-file "${BASH_ENV}" --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && '"export CIRCLE_SHA1='$CIRCLE_SHA1'"' && . ./.circleci/scripts/python_doc_push_script.sh docs/'$target' '$target' site") | docker exec -u jenkins -i "$id" bash) 2>&1'
+
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts
+
+          mkdir -p ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/pytorch.github.io/docs/main ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/pytorch.github.io /tmp/workspace
+
+          # Save the docs build so we can debug any problems
+          export DEBUG_COMMIT_DOCKER_IMAGE=${COMMIT_DOCKER_IMAGE}-debug
+          docker commit "$id" ${DEBUG_COMMIT_DOCKER_IMAGE}
+          time docker push ${DEBUG_COMMIT_DOCKER_IMAGE}
+    - persist_to_workspace:
+        root: /tmp/workspace
+        paths:
+          - .
+    - store_artifacts:
+        path: ~/workspace/build_artifacts/main
+        destination: docs
+
+  pytorch_cpp_doc_build:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-cpp-doc-push
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: large
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc Build and Push
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          # turn v1.12.0rc3 into 1.12
+          tag=$(echo $CIRCLE_TAG | sed -e 's/v*\([0-9]*\.[0-9]*\).*/\1/')
+          target=${tag:-main}
+          echo "building for ${target}"
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --env-file "${BASH_ENV}" --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && '"export CIRCLE_SHA1='$CIRCLE_SHA1'"' && . ./.circleci/scripts/cpp_doc_push_script.sh docs/"$target" main") | docker exec -u jenkins -i "$id" bash) 2>&1'
+
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts
+
+          mkdir -p ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/cppdocs/ /tmp/workspace
+
+          # Save the docs build so we can debug any problems
+          export DEBUG_COMMIT_DOCKER_IMAGE=${COMMIT_DOCKER_IMAGE}-debug
+          docker commit "$id" ${DEBUG_COMMIT_DOCKER_IMAGE}
+          time docker push ${DEBUG_COMMIT_DOCKER_IMAGE}
+
+    - persist_to_workspace:
+        root: /tmp/workspace
+        paths:
+          - .
+
   pytorch_macos_10_15_py3_build:
     environment:
       BUILD_ENVIRONMENT: pytorch-macos-10.15-py3-arm64-build
@@ -512,6 +964,7 @@ jobs:
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export CROSS_COMPILE_ARM64=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
@@ -549,6 +1002,7 @@ jobs:
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             # Install sccache
@@ -570,198 +1024,6 @@ jobs:
           paths:
             - miniconda3
 
-  mac_build:
-    parameters:
-      build-environment:
-        type: string
-        description: Top-level label for what's being built/tested.
-      xcode-version:
-        type: string
-        default: "13.3.1"
-        description: What xcode version to build with.
-      build-generates-artifacts:
-        type: boolean
-        default: true
-        description: if the build generates build artifacts
-      python-version:
-        type: string
-        default: "3.8"
-    macos:
-      xcode: << parameters.xcode-version >>
-    resource_class: medium
-    environment:
-      BUILD_ENVIRONMENT: << parameters.build-environment >>
-      AWS_REGION: us-east-1
-    steps:
-
-      - checkout
-      - run_brew_for_macos_build
-
-      - run:
-          name: Install sccache
-          command: |
-            sudo curl --retry 3 https://s3.amazonaws.com/ossci-macos/sccache_v2.15 --output /usr/local/bin/sccache
-            sudo chmod +x /usr/local/bin/sccache
-            echo "export SCCACHE_BUCKET=ossci-compiler-cache-circleci-v2" >> "${BASH_ENV}"
-            echo "export SCCACHE_S3_KEY_PREFIX=${GITHUB_WORKFLOW}" >> "${BASH_ENV}"
-
-            set +x
-            echo "export AWS_ACCESS_KEY_ID=${CIRCLECI_AWS_ACCESS_KEY_FOR_SCCACHE_S3_BUCKET_V4}" >> "${BASH_ENV}"
-            echo "export AWS_SECRET_ACCESS_KEY=${CIRCLECI_AWS_SECRET_KEY_FOR_SCCACHE_S3_BUCKET_V4}" >> "${BASH_ENV}"
-            set -x
-
-      - run:
-          name: Get workflow job id
-          command: |
-            echo "export OUR_GITHUB_JOB_ID=${CIRCLE_WORKFLOW_JOB_ID}" >> "${BASH_ENV}"
-
-      - run:
-          name: Build
-          command: |
-            set -x
-
-            git submodule sync
-            git submodule update --init --recursive --depth 1 --jobs 0
-
-            export PATH="/usr/local/bin:$PATH"
-            export WORKSPACE_DIR="${HOME}/workspace"
-            mkdir -p "${WORKSPACE_DIR}"
-            MINICONDA_URL="https://repo.anaconda.com/miniconda/Miniconda3-py38_4.12.0-MacOSX-x86_64.sh"
-            if [  << parameters.python-version >> == 3.9.12 ]; then
-              MINICONDA_URL="https://repo.anaconda.com/miniconda/Miniconda3-py39_4.12.0-MacOSX-x86_64.sh"
-            fi
-
-            # If a local installation of conda doesn't exist, we download and install conda
-            if [ ! -d "${WORKSPACE_DIR}/miniconda3" ]; then
-              mkdir -p "${WORKSPACE_DIR}"
-              curl --retry 3 ${MINICONDA_URL} -o "${WORKSPACE_DIR}"/miniconda3.sh
-              bash "${WORKSPACE_DIR}"/miniconda3.sh -b -p "${WORKSPACE_DIR}"/miniconda3
-            fi
-            export PATH="${WORKSPACE_DIR}/miniconda3/bin:$PATH"
-            # shellcheck disable=SC1091
-            source "${WORKSPACE_DIR}"/miniconda3/bin/activate
-
-            brew link --force libomp
-
-            echo "export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname "$(which conda)")/../"}" >> "${BASH_ENV}"
-            .jenkins/pytorch/macos-build.sh
-
-      - when:
-          condition: << parameters.build-generates-artifacts >>
-          steps:
-            - run:
-                name: Archive artifacts into zip
-                command: |
-                  zip -1 -r artifacts.zip dist/ build/.ninja_log build/compile_commands.json .pytorch-test-times.json
-                  cp artifacts.zip /Users/distiller/workspace
-
-      - persist_to_workspace:
-          root: /Users/distiller/workspace/
-          paths:
-            - miniconda3
-            - artifacts.zip
-
-      - store_artifacts:
-          path: /Users/distiller/project/artifacts.zip
-
-  mac_test:
-    parameters:
-      build-environment:
-        type: string
-      shard-number:
-        type: string
-      num-test-shards:
-        type: string
-      xcode-version:
-        type: string
-      test-config:
-        type: string
-        default: 'default'
-
-    macos:
-      xcode: << parameters.xcode-version >>
-    environment:
-      GIT_DEFAULT_BRANCH: 'master'
-      BUILD_ENVIRONMENT: << parameters.build-environment >>
-      TEST_CONFIG: << parameters.test-config >>
-      SHARD_NUMBER: << parameters.shard-number >>
-      NUM_TEST_SHARDS: << parameters.num-test-shards >>
-      PYTORCH_RETRY_TEST_CASES: 1
-      PYTORCH_OVERRIDE_FLAKY_SIGNAL: 1
-    steps:
-      - checkout
-      - attach_workspace:
-          at: ~/workspace
-      - run_brew_for_macos_build
-      - run:
-          name: Test
-          no_output_timeout: "2h"
-          command: |
-            set -x
-
-            git submodule sync --recursive
-            git submodule update --init --recursive
-
-            mv ~/workspace/artifacts.zip .
-            unzip artifacts.zip
-
-            export IN_CI=1
-
-            COMMIT_MESSAGES=$(git cherry -v "origin/${GIT_DEFAULT_BRANCH:-master}")
-
-            export PATH="/usr/local/bin:$PATH"
-            export WORKSPACE_DIR="${HOME}/workspace"
-            mkdir -p "${WORKSPACE_DIR}"
-
-            export PATH="${WORKSPACE_DIR}/miniconda3/bin:$PATH"
-            source "${WORKSPACE_DIR}"/miniconda3/bin/activate
-
-            # sanitize the input commit message and PR body here:
-
-            # trim all new lines from commit messages to avoid issues with batch environment
-            # variable copying. see https://github.com/pytorch/pytorch/pull/80043#issuecomment-1167796028
-            COMMIT_MESSAGES="${COMMIT_MESSAGES//[$'\n\r']}"
-
-            # then trim all special characters like single and double quotes to avoid unescaped inputs to
-            # wreak havoc internally
-            export COMMIT_MESSAGES="${COMMIT_MESSAGES//[\'\"]}"
-
-            python3 -mpip install dist/*.whl
-            .jenkins/pytorch/macos-test.sh
-      - run:
-          name: Copy files for uploading test stats
-          command: |
-            # copy into a parent folder test-reports because we can't use CIRCLEI_BUILD_NUM in path when persisting to workspace
-            mkdir -p test-reports/test-reports_${CIRCLE_BUILD_NUM}/test/test-reports
-            cp -r test/test-reports test-reports/test-reports_${CIRCLE_BUILD_NUM}/test/test-reports
-      - store_test_results:
-          path: test/test-reports
-      - persist_to_workspace:
-          root: /Users/distiller/project/
-          paths:
-            - test-reports
-
-  upload_test_stats:
-    machine: # executor type
-      image: ubuntu-2004:202010-01 # # recommended linux image - includes Ubuntu 20.04, docker 19.03.13, docker-compose 1.27.4
-    steps:
-      - checkout
-      - attach_workspace:
-          at: ~/workspace
-      - run:
-          name: upload
-          command: |
-            set -ex
-            if [ -z ${AWS_ACCESS_KEY_FOR_OSSCI_ARTIFACT_UPLOAD} ]; then
-              echo "No credentials found, cannot upload test stats (are you on a fork?)"
-              exit 0
-            fi
-            cp -r ~/workspace/test-reports/* ~/project
-            pip3 install requests==2.26 rockset==0.8.3 boto3==1.19.12 six==1.16.0
-            export AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_FOR_OSSCI_ARTIFACT_UPLOAD}
-            export AWS_SECRET_ACCESS_KEY=${AWS_SECRET_KEY_FOR_OSSCI_ARTIFACT_UPLOAD}
-            # i dont know how to get the run attempt number for reruns so default to 1
-            python3 -m tools.stats.upload_test_stats --workflow-run-id "${CIRCLE_WORKFLOW_JOB_ID}" --workflow-run-attempt 1 --head-branch << pipeline.git.branch >> --circleci
   pytorch_macos_10_13_py3_test:
     environment:
       BUILD_ENVIRONMENT: pytorch-macos-10.13-py3-test
@@ -777,6 +1039,7 @@ jobs:
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             chmod a+x .jenkins/pytorch/macos-test.sh
@@ -789,6 +1052,7 @@ jobs:
             source /Users/distiller/workspace/miniconda3/bin/activate
             python3 -m pip install boto3==1.19.12
 
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             # Using the same IAM user to write stats to our OSS bucket
@@ -814,6 +1078,7 @@ jobs:
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export BUILD_LITE_INTERPRETER=1
             export JOB_BASE_NAME=$CIRCLE_JOB
             chmod a+x ${HOME}/project/.jenkins/pytorch/macos-lite-interpreter-build-test.sh
@@ -903,6 +1168,9 @@ jobs:
           output_image=$docker_image_libtorch_android_x86_32-gradle
           docker commit "$id_x86_32" ${output_image}
           time docker push ${output_image}
+    - upload_binary_size_for_android_build:
+        build_type: prebuilt
+        artifacts: /home/circleci/workspace/build_android_artifacts/artifacts.tgz
     - store_artifacts:
         path: ~/workspace/build_android_artifacts/artifacts.tgz
         destination: artifacts.tgz
@@ -978,6 +1246,9 @@ jobs:
           output_image=${docker_image_libtorch_android_x86_32}-gradle
           docker commit "$id" ${output_image}
           time docker push ${output_image}
+    - upload_binary_size_for_android_build:
+        build_type: prebuilt-single
+        artifacts: /home/circleci/workspace/build_android_x86_32_artifacts/artifacts.tgz
     - store_artifacts:
         path: ~/workspace/build_android_x86_32_artifacts/artifacts.tgz
         destination: artifacts.tgz
@@ -987,43 +1258,10 @@ jobs:
     macos:
       xcode: "12.5.1"
     steps:
-      - run:
-          name: checkout with retry
-          command: |
-            checkout() {
-              set -ex
-              # Workaround old docker images with incorrect $HOME
-              # check https://github.com/docker/docker/issues/2968 for details
-              if [ "${HOME}" = "/" ]
-                then
-                export HOME=$(getent passwd $(id -un) | cut -d: -f6)
-              fi
-
-              mkdir -p ~/.ssh
-
-              echo 'github.com ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEAq2A7hRGmdnm9tUDbO9IDSwBK6TbQa+PXYPCPy6rbTrTtw7PHkccKrpp0yVhp5HdEIcKr6pLlVDBfOLX9QUsyCOV0wzfjIJNlGEYsdlLJizHhbn2mUjvSAHQqZETYP81eFzLQNnPHt4EVVUh7VfDESU84KezmD5QlWpXLmvU31/yMf+Se8xhHTvKSCZIFImWwoG6mbUoWf9nzpIoaSjB+weqqUUmpaaasXVal72J+UX2B+2RPW3RcT0eOzQgqlJL3RKrTJvdsjE3JEAvGq3lGHSZXy28G3skua2SmVi/w4yCE6gbODqnTWlg7+wC604ydGXA8VJiS5ap43JXiUFFAaQ==
-              ' >> ~/.ssh/known_hosts
-
-              # use git+ssh instead of https
-              git config --global url."ssh://git@github.com".insteadOf "https://github.com" || true
-              git config --global gc.auto 0 || true
-
-              echo 'Cloning git repository'
-              mkdir -p '/Users/distiller/project'
-              cd '/Users/distiller/project'
-              git clone "$CIRCLE_REPOSITORY_URL" .
-              echo 'Checking out branch'
-              git checkout --force -B "$CIRCLE_BRANCH" "$CIRCLE_SHA1"
-              git --no-pager log --no-color -n 1 --format='HEAD is now at %h %s'
-            }
-
-            retry () {
-              $* || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
-            }
-            retry checkout
+      - checkout
       - run_brew_for_ios_build
       - run:
-          name: Setup Fastlane
+          name: Run Fastlane
           no_output_timeout: "1h"
           command: |
             set -e
@@ -1031,11 +1269,26 @@ jobs:
             cd ${PROJ_ROOT}/ios/TestApp
             # install fastlane
             sudo gem install bundler && bundle install
+            # install certificates
+            echo ${IOS_CERT_KEY_2022} >> cert.txt
+            base64 --decode cert.txt -o Certificates.p12
+            rm cert.txt
+            bundle exec fastlane install_root_cert
+            bundle exec fastlane install_dev_cert
+            # install the provisioning profile
+            PROFILE=PyTorch_CI_2022.mobileprovision
+            PROVISIONING_PROFILES=~/Library/MobileDevice/Provisioning\ Profiles
+            mkdir -pv "${PROVISIONING_PROFILES}"
+            cd "${PROVISIONING_PROFILES}"
+            echo ${IOS_SIGN_KEY_2022} >> cert.txt
+            base64 --decode cert.txt -o ${PROFILE}
+            rm cert.txt
       - run:
           name: Build
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             WORKSPACE=/Users/distiller/workspace
             PROJ_ROOT=/Users/distiller/project
             export TCLLIBPATH="/usr/local/lib"
@@ -1088,12 +1341,18 @@ jobs:
           command: |
             set -e
             PROJ_ROOT=/Users/distiller/project
+            PROFILE=PyTorch_CI_2022
             # run the ruby build script
             if ! [ -x "$(command -v xcodebuild)" ]; then
               echo 'Error: xcodebuild is not installed.'
               exit 1
             fi
-            ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM}
+            echo ${IOS_DEV_TEAM_ID}
+            if [ ${IOS_PLATFORM} != "SIMULATOR" ]; then
+              ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM} -c ${PROFILE} -t ${IOS_DEV_TEAM_ID}
+            else
+              ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM}
+            fi
             if ! [ "$?" -eq "0" ]; then
               echo 'xcodebuild failed!'
               exit 1
@@ -1116,13 +1375,12 @@ jobs:
             cd ${PROJ_ROOT}/ios/TestApp/benchmark
             mkdir -p ../models
             if [ ${USE_COREML_DELEGATE} == 1 ]; then
-              pip install coremltools==5.0b5 protobuf==3.20.1 six==1.16.0
+              pip install coremltools==5.0b5
+              pip install six
               python coreml_backend.py
             else
-              cd "${PROJ_ROOT}"
-              python test/mobile/model_test/gen_test_model.py ios-test
+              python trace_model.py
             fi
-            cd "${PROJ_ROOT}/ios/TestApp/benchmark"
             if [ ${BUILD_LITE_INTERPRETER} == 1 ]; then
               echo "Setting up the TestApp for LiteInterpreter"
               ruby setup.rb --lite 1
@@ -1130,10 +1388,10 @@ jobs:
               echo "Setting up the TestApp for Full JIT"
               ruby setup.rb
             fi
-            cd "${PROJ_ROOT}/ios/TestApp"
-            # instruments -s -devices
-            if [ "${BUILD_LITE_INTERPRETER}" == 1 ]; then
-              if [ "${USE_COREML_DELEGATE}" == 1 ]; then
+            cd ${PROJ_ROOT}/ios/TestApp
+            instruments -s -devices
+            if [ ${BUILD_LITE_INTERPRETER} == 1 ]; then
+              if [ ${USE_COREML_DELEGATE} == 1 ]; then
                 fastlane scan --only_testing TestAppTests/TestAppTests/testCoreML
               else
                 fastlane scan --only_testing TestAppTests/TestAppTests/testLiteInterpreter
@@ -1236,6 +1494,30 @@ jobs:
             python3 -m pip install requests
             python3 ./.circleci/scripts/trigger_azure_pipeline.py
 
+  pytorch_doc_test:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-doc-test
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: medium
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc test
+        no_output_timeout: "30m"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && . ./.jenkins/pytorch/docs-test.sh") | docker exec -u jenkins -i "$id" bash) 2>&1'
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts
+
   # update_s3_htmls job
   # These jobs create html files for every cpu/cu## folder in s3. The html
   # files just store the names of all the files in that folder (which are
@@ -1447,107 +1729,4 @@ workflows:
             branches:
               only:
                 - postnightly
-      - mac_build:
-          name: macos-12-py3-x86-64-build
-          build-environment: macos-12-py3-x86-64
-          xcode-version: "13.3.1"
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - mac_test:
-          name: macos-12-py3-x86-64-test-1-2-default
-          build-environment: macos-12-py3-x86-64
-          xcode-version: "13.3.1"
-          shard-number: "1"
-          num-test-shards: "2"
-          requires:
-            - macos-12-py3-x86-64-build
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - mac_test:
-          name: macos-12-py3-x86-64-test-2-2-default
-          build-environment: macos-12-py3-x86-64
-          xcode-version: "13.3.1"
-          shard-number: "2"
-          num-test-shards: "2"
-          requires:
-            - macos-12-py3-x86-64-build
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - mac_test:
-          name: macos-12-py3-x86-64-test-1-1-functorch
-          build-environment: macos-12-py3-x86-64
-          xcode-version: "13.3.1"
-          shard-number: "1"
-          num-test-shards: "1"
-          test-config: functorch
-          requires:
-            - macos-12-py3-x86-64-build
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - mac_build:
-          name: macos-12-py3-x86-64-lite-interpreter-build-test
-          build-environment: macos-12-py3-lite-interpreter-x86-64
-          xcode-version: "13.3.1"
-          build-generates-artifacts: false
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - mac_build:
-          name: macos-12-py3-arm64-build
-          build-environment: macos-12-py3-arm64
-          xcode-version: "13.3.1"
-          python-version: "3.9.12"
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - upload_test_stats:
-          name: upload test status
-          requires:
-            - macos-12-py3-x86-64-test-1-2-default
-            - macos-12-py3-x86-64-test-2-2-default
-            - macos-12-py3-x86-64-test-1-1-functorch
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-      - pytorch_ios_build:
-          build_environment: ios-12-5-1-x86-64
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-          ios_arch: x86_64
-          ios_platform: SIMULATOR
-          lite_interpreter: "1"
-          name: ios-12-5-1-x86-64
-      - pytorch_ios_build:
-          build_environment: ios-12-5-1-x86-64-coreml
-          filters:
-            branches:
-              ignore:
-                - nightly
-                - postnightly
-          ios_arch: x86_64
-          ios_platform: SIMULATOR
-          lite_interpreter: "1"
-          name: ios-12-5-1-x86-64-coreml
-          use_coreml: "1"
     when: << pipeline.parameters.run_build >>

.circleci/docker/android/build.gradle

@@ -53,7 +53,7 @@ dependencies {
     implementation 'androidx.appcompat:appcompat:1.0.0'
     implementation 'com.facebook.fbjni:fbjni-java-only:0.2.2'
     implementation 'com.google.code.findbugs:jsr305:3.0.1'
-    implementation 'com.facebook.soloader:nativeloader:0.10.4'
+    implementation 'com.facebook.soloader:nativeloader:0.10.1'
 
     implementation 'junit:junit:' + rootProject.junitVersion
     implementation 'androidx.test:core:' + rootProject.coreVersion

.circleci/docker/build.sh

@@ -54,8 +54,6 @@ elif [[ "$image" == *-bionic* ]]; then
   UBUNTU_VERSION=18.04
 elif [[ "$image" == *-focal* ]]; then
   UBUNTU_VERSION=20.04
-elif [[ "$image" == *-jammy* ]]; then
-  UBUNTU_VERSION=22.04
 elif [[ "$image" == *ubuntu* ]]; then
   extract_version_from_image_name ubuntu UBUNTU_VERSION
 elif [[ "$image" == *centos* ]]; then
@@ -72,8 +70,7 @@ else
 fi
 
 DOCKERFILE="${OS}/Dockerfile"
-# When using ubuntu - 22.04, start from Ubuntu docker image, instead of nvidia/cuda docker image.
-if [[ "$image" == *cuda* && "$UBUNTU_VERSION" != "22.04" ]]; then
+if [[ "$image" == *cuda* ]]; then
   DOCKERFILE="${OS}-cuda/Dockerfile"
 elif [[ "$image" == *rocm* ]]; then
   DOCKERFILE="${OS}-rocm/Dockerfile"
@@ -84,8 +81,6 @@ if [[ "$image" == *xenial* ]] || [[ "$image" == *bionic* ]]; then
 fi
 
 TRAVIS_DL_URL_PREFIX="https://s3.amazonaws.com/travis-python-archives/binaries/ubuntu/14.04/x86_64"
-_UCX_COMMIT=31e74cac7bee0ef66bef2af72e7d86d9c282e5ab
-_UCC_COMMIT=12944da33f911daf505d9bbc51411233d0ed85e1
 
 # It's annoying to rename jobs every time you want to rewrite a
 # configuration, so we hardcode everything here rather than do it
@@ -96,6 +91,14 @@ case "$image" in
     GCC_VERSION=7
     # Do not install PROTOBUF, DB, and VISION as a test
     ;;
+  pytorch-linux-xenial-py3.7-gcc5.4)
+    ANACONDA_PYTHON_VERSION=3.7
+    GCC_VERSION=5
+    PROTOBUF=yes
+    DB=yes
+    VISION=yes
+    KATEX=yes
+    ;;
   pytorch-linux-xenial-py3.7-gcc7.2)
     ANACONDA_PYTHON_VERSION=3.7
     GCC_VERSION=7
@@ -141,28 +144,14 @@ case "$image" in
     KATEX=yes
     ;;
   pytorch-linux-bionic-cuda11.6-cudnn8-py3-gcc7)
-    CUDA_VERSION=11.6.2
+    CUDA_VERSION=11.6.0
     CUDNN_VERSION=8
-    ANACONDA_PYTHON_VERSION=3.10
+    ANACONDA_PYTHON_VERSION=3.7
     GCC_VERSION=7
     PROTOBUF=yes
     DB=yes
     VISION=yes
     KATEX=yes
-    UCX_COMMIT=${_UCX_COMMIT}
-    UCC_COMMIT=${_UCC_COMMIT}
-    ;;
-  pytorch-linux-bionic-cuda11.7-cudnn8-py3-gcc7)
-    CUDA_VERSION=11.7.0
-    CUDNN_VERSION=8
-    ANACONDA_PYTHON_VERSION=3.10
-    GCC_VERSION=7
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    KATEX=yes
-    UCX_COMMIT=${_UCX_COMMIT}
-    UCC_COMMIT=${_UCC_COMMIT}
     ;;
   pytorch-linux-xenial-py3-clang5-asan)
     ANACONDA_PYTHON_VERSION=3.7
@@ -178,13 +167,6 @@ case "$image" in
     DB=yes
     VISION=yes
     ;;
-  pytorch-linux-focal-py3-clang7-asan)
-    ANACONDA_PYTHON_VERSION=3.7
-    CLANG_VERSION=7
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    ;;
   pytorch-linux-xenial-py3-clang7-onnx)
     ANACONDA_PYTHON_VERSION=3.7
     CLANG_VERSION=7
@@ -192,13 +174,6 @@ case "$image" in
     DB=yes
     VISION=yes
     ;;
-  pytorch-linux-focal-py3-clang10-onnx)
-    ANACONDA_PYTHON_VERSION=3.7
-    CLANG_VERSION=10
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    ;;
   pytorch-linux-xenial-py3-clang5-android-ndk-r19c)
     ANACONDA_PYTHON_VERSION=3.7
     CLANG_VERSION=5.0
@@ -250,7 +225,15 @@ case "$image" in
     DB=yes
     VISION=yes
     ;;
-  pytorch-linux-focal-rocm5.1-py3.7)
+  pytorch-linux-bionic-rocm5.0-py3.7)
+    ANACONDA_PYTHON_VERSION=3.7
+    GCC_VERSION=9
+    PROTOBUF=yes
+    DB=yes
+    VISION=yes
+    ROCM_VERSION=5.0
+    ;;
+  pytorch-linux-bionic-rocm5.1-py3.7)
     ANACONDA_PYTHON_VERSION=3.7
     GCC_VERSION=9
     PROTOBUF=yes
@@ -258,41 +241,15 @@ case "$image" in
     VISION=yes
     ROCM_VERSION=5.1.1
     ;;
-  pytorch-linux-focal-rocm5.2-py3.7)
-    ANACONDA_PYTHON_VERSION=3.7
-    GCC_VERSION=9
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    ROCM_VERSION=5.2
-    ;;
   pytorch-linux-focal-py3.7-gcc7)
     ANACONDA_PYTHON_VERSION=3.7
-    CMAKE_VERSION=3.16.9  # Required for precompiled header support
+    CMAKE_VERSION=3.12.4  # To make sure XNNPACK is enabled for the BACKWARDS_COMPAT_TEST used with this image
     GCC_VERSION=7
     PROTOBUF=yes
     DB=yes
     VISION=yes
     KATEX=yes
     ;;
-  pytorch-linux-jammy-cuda11.6-cudnn8-py3.8-clang12)
-    ANACONDA_PYTHON_VERSION=3.8
-    CUDA_VERSION=11.6
-    CUDNN_VERSION=8
-    CLANG_VERSION=12
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    ;;
-  pytorch-linux-jammy-cuda11.7-cudnn8-py3.8-clang12)
-    ANACONDA_PYTHON_VERSION=3.8
-    CUDA_VERSION=11.7
-    CUDNN_VERSION=8
-    CLANG_VERSION=12
-    PROTOBUF=yes
-    DB=yes
-    VISION=yes
-    ;;
   *)
     # Catch-all for builds that are not hardcoded.
     PROTOBUF=yes
@@ -379,10 +336,8 @@ docker build \
        --build-arg "NINJA_VERSION=${NINJA_VERSION:-}" \
        --build-arg "KATEX=${KATEX:-}" \
        --build-arg "ROCM_VERSION=${ROCM_VERSION:-}" \
-       --build-arg "PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH:-gfx906}" \
+       --build-arg "PYTORCH_ROCM_ARCH=${PYTORCH_ROCM_ARCH:-gfx900;gfx906}" \
        --build-arg "IMAGE_NAME=${IMAGE_NAME}" \
-       --build-arg "UCX_COMMIT=${UCX_COMMIT}" \
-       --build-arg "UCC_COMMIT=${UCC_COMMIT}" \
        -f $(dirname ${DOCKERFILE})/Dockerfile \
        -t "$tmp_tag" \
        "$@" \

.circleci/docker/build_docker.sh

@@ -18,7 +18,6 @@ tag="${DOCKER_TAG}"
 
 registry="308535385114.dkr.ecr.us-east-1.amazonaws.com"
 image="${registry}/pytorch/${IMAGE_NAME}"
-ghcr_image="ghcr.io/pytorch/ci-image"
 
 login() {
   aws ecr get-authorization-token --region us-east-1 --output text --query 'authorizationData[].authorizationToken' |
@@ -47,22 +46,7 @@ fi
 # Build new image
 ./build.sh ${IMAGE_NAME} -t "${image}:${tag}"
 
-# Only push if `DOCKER_SKIP_PUSH` = false
-if [ "${DOCKER_SKIP_PUSH:-true}" = "false" ]; then
-  # Only push if docker image doesn't exist already.
-  # ECR image tags are immutable so this will avoid pushing if only just testing if the docker jobs work
-  # NOTE: The only workflow that should push these images should be the docker-builds.yml workflow
-  if ! docker manifest inspect "${image}:${tag}" >/dev/null 2>/dev/null; then
-    docker push "${image}:${tag}"
-  fi
-
-  if [ "${PUSH_GHCR_IMAGE:-}" = "true" ]; then
-    # Push docker image to the ghcr.io
-    echo $GHCR_PAT | docker login ghcr.io -u pytorch --password-stdin
-    docker tag "${image}:${tag}" "${ghcr_image}:${IMAGE_NAME}-${tag}"
-    docker push "${ghcr_image}:${IMAGE_NAME}-${tag}"
-  fi
-fi
+docker push "${image}:${tag}"
 
 if [ -z "${DOCKER_SKIP_S3_UPLOAD:-}" ]; then
   trap "rm -rf ${IMAGE_NAME}:${tag}.tar" EXIT

.circleci/docker/centos-rocm/Dockerfile

@@ -12,7 +12,7 @@ ENV PYTORCH_ROCM_ARCH ${PYTORCH_ROCM_ARCH}
 
 # Install common dependencies (so that this step can be cached separately)
 ARG EC2
-COPY ./common/install_base.sh install_base.sh
+ADD ./common/install_base.sh install_base.sh
 RUN bash ./install_base.sh && rm install_base.sh
 
 # Update CentOS git version
@@ -23,57 +23,54 @@ RUN yum install -y git
 
 # Install devtoolset
 ARG DEVTOOLSET_VERSION
-COPY ./common/install_devtoolset.sh install_devtoolset.sh
+ADD ./common/install_devtoolset.sh install_devtoolset.sh
 RUN bash ./install_devtoolset.sh && rm install_devtoolset.sh
 ENV BASH_ENV "/etc/profile"
 
 # (optional) Install non-default glibc version
 ARG GLIBC_VERSION
-COPY ./common/install_glibc.sh install_glibc.sh
+ADD ./common/install_glibc.sh install_glibc.sh
 RUN if [ -n "${GLIBC_VERSION}" ]; then bash ./install_glibc.sh; fi
 RUN rm install_glibc.sh
 
 # Install user
-COPY ./common/install_user.sh install_user.sh
+ADD ./common/install_user.sh install_user.sh
 RUN bash ./install_user.sh && rm install_user.sh
 
 # Install conda and other packages (e.g., numpy, pytest)
 ENV PATH /opt/conda/bin:$PATH
 ARG ANACONDA_PYTHON_VERSION
-COPY requirements-ci.txt /opt/conda/requirements-ci.txt
-COPY ./common/install_conda.sh install_conda.sh
+ADD requirements-ci.txt /opt/conda/requirements-ci.txt
+ADD ./common/install_conda.sh install_conda.sh
 RUN bash ./install_conda.sh && rm install_conda.sh
 RUN rm /opt/conda/requirements-ci.txt
 
 # (optional) Install protobuf for ONNX
 ARG PROTOBUF
-COPY ./common/install_protobuf.sh install_protobuf.sh
+ADD ./common/install_protobuf.sh install_protobuf.sh
 RUN if [ -n "${PROTOBUF}" ]; then bash ./install_protobuf.sh; fi
 RUN rm install_protobuf.sh
 ENV INSTALLED_PROTOBUF ${PROTOBUF}
 
 # (optional) Install database packages like LMDB and LevelDB
 ARG DB
-COPY ./common/install_db.sh install_db.sh
+ADD ./common/install_db.sh install_db.sh
 RUN if [ -n "${DB}" ]; then bash ./install_db.sh; fi
 RUN rm install_db.sh
 ENV INSTALLED_DB ${DB}
 
 # (optional) Install vision packages like OpenCV and ffmpeg
 ARG VISION
-COPY ./common/install_vision.sh install_vision.sh
+ADD ./common/install_vision.sh install_vision.sh
 RUN if [ -n "${VISION}" ]; then bash ./install_vision.sh; fi
 RUN rm install_vision.sh
 ENV INSTALLED_VISION ${VISION}
 
 # Install rocm
 ARG ROCM_VERSION
-COPY ./common/install_rocm.sh install_rocm.sh
+ADD ./common/install_rocm.sh install_rocm.sh
 RUN bash ./install_rocm.sh
 RUN rm install_rocm.sh
-COPY ./common/install_rocm_magma.sh install_rocm_magma.sh
-RUN bash ./install_rocm_magma.sh
-RUN rm install_rocm_magma.sh
 ENV PATH /opt/rocm/bin:$PATH
 ENV PATH /opt/rocm/hcc/bin:$PATH
 ENV PATH /opt/rocm/hip/bin:$PATH
@@ -85,18 +82,18 @@ ENV LC_ALL en_US.utf8
 
 # (optional) Install non-default CMake version
 ARG CMAKE_VERSION
-COPY ./common/install_cmake.sh install_cmake.sh
+ADD ./common/install_cmake.sh install_cmake.sh
 RUN if [ -n "${CMAKE_VERSION}" ]; then bash ./install_cmake.sh; fi
 RUN rm install_cmake.sh
 
 # (optional) Install non-default Ninja version
 ARG NINJA_VERSION
-COPY ./common/install_ninja.sh install_ninja.sh
+ADD ./common/install_ninja.sh install_ninja.sh
 RUN if [ -n "${NINJA_VERSION}" ]; then bash ./install_ninja.sh; fi
 RUN rm install_ninja.sh
 
 # Install ccache/sccache (do this last, so we get priority in PATH)
-COPY ./common/install_cache.sh install_cache.sh
+ADD ./common/install_cache.sh install_cache.sh
 ENV PATH /opt/cache/bin:$PATH
 RUN bash ./install_cache.sh && rm install_cache.sh
 

.circleci/docker/common/install_base.sh

@@ -15,22 +15,11 @@ install_ubuntu() {
   elif [[ "$UBUNTU_VERSION" == "20.04"* ]]; then
     cmake3="cmake=3.16*"
     maybe_libiomp_dev=""
-  elif [[ "$UBUNTU_VERSION" == "22.04"* ]]; then
-    cmake3="cmake=3.22*"
-    maybe_libiomp_dev=""
   else
     cmake3="cmake=3.5*"
     maybe_libiomp_dev="libiomp-dev"
   fi
 
-  if [[ "$CLANG_VERSION" == 12 ]]; then
-    maybe_libomp_dev="libomp-12-dev"
-  elif [[ "$CLANG_VERSION" == 10 ]]; then
-    maybe_libomp_dev="libomp-10-dev"
-  else
-    maybe_libomp_dev=""
-  fi
-
   # TODO: Remove this once nvidia package repos are back online
   # Comment out nvidia repositories to prevent them from getting apt-get updated, see https://github.com/pytorch/pytorch/issues/74968
   # shellcheck disable=SC2046
@@ -40,12 +29,10 @@ install_ubuntu() {
   apt-get update
   # TODO: Some of these may not be necessary
   ccache_deps="asciidoc docbook-xml docbook-xsl xsltproc"
-  deploy_deps="libffi-dev libbz2-dev libreadline-dev libncurses5-dev libncursesw5-dev libgdbm-dev libsqlite3-dev uuid-dev tk-dev"
   numpy_deps="gfortran"
   apt-get install -y --no-install-recommends \
     $ccache_deps \
     $numpy_deps \
-    ${deploy_deps} \
     ${cmake3} \
     apt-transport-https \
     autoconf \
@@ -62,32 +49,15 @@ install_ubuntu() {
     libjpeg-dev \
     libasound2-dev \
     libsndfile-dev \
-    ${maybe_libomp_dev} \
     software-properties-common \
     wget \
     sudo \
-    vim \
-    jq \
-    libtool
+    vim
 
   # Should resolve issues related to various apt package repository cert issues
   # see: https://github.com/pytorch/pytorch/issues/65931
   apt-get install -y libgnutls30
 
-  # cuda-toolkit does not work with gcc-11.2.0 which is default in Ubunutu 22.04
-  # see: https://github.com/NVlabs/instant-ngp/issues/119
-  if [[ "$UBUNTU_VERSION" == "22.04"* ]]; then
-    apt-get install -y g++-10
-    update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-10 30
-    update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-10 30
-    update-alternatives --install /usr/bin/gcov gcov /usr/bin/gcov-10 30
-
-    # https://www.spinics.net/lists/libreoffice/msg07549.html
-    sudo rm -rf /usr/lib/gcc/x86_64-linux-gnu/11
-    wget https://github.com/gcc-mirror/gcc/commit/2b2d97fc545635a0f6aa9c9ee3b017394bc494bf.patch -O noexecpt.patch
-    sudo patch  /usr/include/c++/10/bits/range_access.h noexecpt.patch
-  fi
-
   # Cleanup package manager
   apt-get autoclean && apt-get clean
   rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*

.circleci/docker/common/install_cache.sh

@@ -5,9 +5,7 @@ set -ex
 install_ubuntu() {
   echo "Preparing to build sccache from source"
   apt-get update
-  # libssl-dev will not work as it is upgraded to libssl3 in Ubuntu-22.04.
-  # Instead use lib and headers from OpenSSL1.1 installed in `install_openssl.sh``
-  apt-get install -y cargo
+  apt-get install -y cargo pkg-config libssl-dev
   echo "Checking out sccache repo"
   git clone https://github.com/pytorch/sccache
   cd sccache
@@ -48,9 +46,7 @@ fi
 chmod a+x /opt/cache/bin/sccache
 
 function write_sccache_stub() {
-  # Unset LD_PRELOAD for ps because of asan + ps issues
-  # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90589
-  printf "#!/bin/sh\nif [ \$(env -u LD_PRELOAD ps -p \$PPID -o comm=) != sccache ]; then\n  exec sccache $(which $1) \"\$@\"\nelse\n  exec $(which $1) \"\$@\"\nfi" > "/opt/cache/bin/$1"
+  printf "#!/bin/sh\nif [ \$(ps -p \$PPID -o comm=) != sccache ]; then\n  exec sccache $(which $1) \"\$@\"\nelse\n  exec $(which $1) \"\$@\"\nfi" > "/opt/cache/bin/$1"
   chmod a+x "/opt/cache/bin/$1"
 }
 

.circleci/docker/common/install_clang.sh

@@ -13,9 +13,6 @@ if [ -n "$CLANG_VERSION" ]; then
     sudo apt-get install  -y --no-install-recommends gpg-agent
     wget --no-check-certificate -O - https://apt.llvm.org/llvm-snapshot.gpg.key | sudo apt-key add  -
     apt-add-repository "deb http://apt.llvm.org/bionic/ llvm-toolchain-bionic-${CLANG_VERSION} main"
-  elif [[ $UBUNTU_VERSION == 22.04 ]]; then
-    # work around ubuntu apt-get conflicts
-    sudo apt-get -y -f install
   fi
 
   sudo apt-get update

.circleci/docker/common/install_conda.sh

@@ -55,10 +55,8 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
   # Ensure we run conda in a directory that jenkins has write access to
   pushd /opt/conda
 
-  # Prevent conda from updating to 4.14.0, which causes docker build failures
-  # See https://hud.pytorch.org/pytorch/pytorch/commit/754d7f05b6841e555cea5a4b2c505dd9e0baec1d
-  # Uncomment the below when resolved to track the latest conda update
-  # as_jenkins conda update -y -n base conda
+  # Track latest conda update
+  as_jenkins conda update -y -n base conda
 
   # Install correct Python version
   as_jenkins conda install -y python="$ANACONDA_PYTHON_VERSION"
@@ -75,21 +73,19 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
   }
 
   # Install PyTorch conda deps, as per https://github.com/pytorch/pytorch README
-  # DO NOT install cmake here as it would install a version newer than 3.13, but
-  # we want to pin to version 3.13.
-  CONDA_COMMON_DEPS="astunparse pyyaml mkl=2022.0.1 mkl-include=2022.0.1 setuptools cffi future six"
-  if [ "$ANACONDA_PYTHON_VERSION" = "3.10" ]; then
+  # DO NOT install cmake here as it would install a version newer than 3.10, but
+  # we want to pin to version 3.10.
+  if [ "$ANACONDA_PYTHON_VERSION" = "3.9" ]; then
     # Install llvm-8 as it is required to compile llvmlite-0.30.0 from source
-    conda_install numpy=1.21.2 ${CONDA_COMMON_DEPS} llvmdev=8.0.0
-  elif [ "$ANACONDA_PYTHON_VERSION" = "3.9" ]; then
-    # Install llvm-8 as it is required to compile llvmlite-0.30.0 from source
-    conda_install numpy=1.19.2 ${CONDA_COMMON_DEPS} llvmdev=8.0.0
+    conda_install numpy=1.19.2 astunparse pyyaml mkl mkl-include setuptools cffi future six llvmdev=8.0.0
   elif [ "$ANACONDA_PYTHON_VERSION" = "3.8" ]; then
     # Install llvm-8 as it is required to compile llvmlite-0.30.0 from source
-    conda_install numpy=1.18.5 ${CONDA_COMMON_DEPS} llvmdev=8.0.0
+    conda_install numpy=1.18.5 astunparse pyyaml mkl mkl-include setuptools cffi future six llvmdev=8.0.0
+  elif [ "$ANACONDA_PYTHON_VERSION" = "3.7" ]; then
+    # DO NOT install dataclasses if installing python-3.7, since its part of python-3.7 core packages
+    conda_install numpy=1.18.5 astunparse pyyaml mkl mkl-include setuptools cffi future six typing_extensions
   else
-    # Install `typing_extensions` for 3.7
-    conda_install numpy=1.18.5 ${CONDA_COMMON_DEPS} typing_extensions
+    conda_install numpy=1.18.5 astunparse pyyaml mkl mkl-include setuptools cffi future six dataclasses typing_extensions
   fi
 
   # Magma package names are concatenation of CUDA major and minor ignoring revision

.circleci/docker/common/install_cudnn.sh

@@ -4,13 +4,7 @@ if [[ ${CUDNN_VERSION} == 8 ]]; then
     # cuDNN license: https://developer.nvidia.com/cudnn/license_agreement
     mkdir tmp_cudnn && cd tmp_cudnn
     CUDNN_NAME="cudnn-linux-x86_64-8.3.2.44_cuda11.5-archive"
-    if [[ ${CUDA_VERSION:0:4} == "11.7" ]]; then
-        CUDNN_NAME="cudnn-linux-x86_64-8.5.0.96_cuda11-archive"
-        curl -OLs https://ossci-linux.s3.amazonaws.com/${CUDNN_NAME}.tar.xz
-    else
-        curl -OLs  https://developer.download.nvidia.com/compute/redist/cudnn/v8.3.2/local_installers/11.5/${CUDNN_NAME}.tar.xz
-    fi
-
+    curl -OLs  https://developer.download.nvidia.com/compute/redist/cudnn/v8.3.2/local_installers/11.5/${CUDNN_NAME}.tar.xz
     tar xf ${CUDNN_NAME}.tar.xz
     cp -a ${CUDNN_NAME}/include/* /usr/include/
     cp -a ${CUDNN_NAME}/include/* /usr/local/cuda/include/

.circleci/docker/common/install_katex.sh

@@ -17,8 +17,6 @@ if [ -n "$KATEX" ]; then
   apt-get install -y --no-install-recommends yarn
   yarn global add katex --prefix /usr/local
 
-  sudo apt-get -y install doxygen
-
   apt-get autoclean && apt-get clean
   rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
 

.circleci/docker/common/install_openssl.sh

@@ -10,7 +10,5 @@ cd "${OPENSSL}"
 ./config --prefix=/opt/openssl -d '-Wl,--enable-new-dtags,-rpath,$(LIBRPATH)'
 # NOTE: openssl install errors out when built with the -j option
 make -j6; make install_sw
-# Link the ssl libraries to the /usr/lib folder.
-sudo ln -s /opt/openssl/lib/lib* /usr/lib
 cd ..
 rm -rf "${OPENSSL}"

.circleci/docker/common/install_rocm.sh

@@ -2,12 +2,40 @@
 
 set -ex
 
+install_magma() {
+    # "install" hipMAGMA into /opt/rocm/magma by copying after build
+    git clone https://bitbucket.org/icl/magma.git
+    pushd magma
+    # Fixes memory leaks of magma found while executing linalg UTs
+    git checkout 5959b8783e45f1809812ed96ae762f38ee701972
+    cp make.inc-examples/make.inc.hip-gcc-mkl make.inc
+    echo 'LIBDIR += -L$(MKLROOT)/lib' >> make.inc
+    echo 'LIB += -Wl,--enable-new-dtags -Wl,--rpath,/opt/rocm/lib -Wl,--rpath,$(MKLROOT)/lib -Wl,--rpath,/opt/rocm/magma/lib' >> make.inc
+    echo 'DEVCCFLAGS += --gpu-max-threads-per-block=256' >> make.inc
+    export PATH="${PATH}:/opt/rocm/bin"
+    if [[ -n "$PYTORCH_ROCM_ARCH" ]]; then
+      amdgpu_targets=`echo $PYTORCH_ROCM_ARCH | sed 's/;/ /g'`
+    else
+      amdgpu_targets=`rocm_agent_enumerator | grep -v gfx000 | sort -u | xargs`
+    fi
+    for arch in $amdgpu_targets; do
+      echo "DEVCCFLAGS += --amdgpu-target=$arch" >> make.inc
+    done
+    # hipcc with openmp flag may cause isnan() on __device__ not to be found; depending on context, compiler may attempt to match with host definition
+    sed -i 's/^FOPENMP/#FOPENMP/g' make.inc
+    make -f make.gen.hipMAGMA -j $(nproc)
+    LANG=C.UTF-8 make lib/libmagma.so -j $(nproc) MKLROOT=/opt/conda
+    make testing/testing_dgemm -j $(nproc) MKLROOT=/opt/conda
+    popd
+    mv magma /opt/rocm
+}
+
 ver() {
     printf "%3d%03d%03d%03d" $(echo "$1" | tr '.' ' ');
 }
 
 # Map ROCm version to AMDGPU version
-declare -A AMDGPU_VERSIONS=( ["5.0"]="21.50" ["5.1.1"]="22.10.1" ["5.2"]="22.20" )
+declare -A AMDGPU_VERSIONS=( ["4.5.2"]="21.40.2" ["5.0"]="21.50" ["5.1.1"]="22.10.1" )
 
 install_ubuntu() {
     apt-get update
@@ -61,6 +89,8 @@ install_ubuntu() {
       DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-unauthenticated ${MIOPENKERNELS}
     fi
 
+    install_magma
+
     # Cleanup
     apt-get autoclean && apt-get clean
     rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
@@ -105,6 +135,8 @@ install_centos() {
                    rocprofiler-dev \
                    roctracer-dev
 
+  install_magma
+
   # Cleanup
   yum clean all
   rm -rf /var/cache/yum

.circleci/docker/common/install_rocm_magma.sh

@@ -1,29 +0,0 @@
-#!/bin/bash
-
-set -ex
-
-# "install" hipMAGMA into /opt/rocm/magma by copying after build
-git clone https://bitbucket.org/icl/magma.git
-pushd magma
-# Fixes memory leaks of magma found while executing linalg UTs
-git checkout 5959b8783e45f1809812ed96ae762f38ee701972
-cp make.inc-examples/make.inc.hip-gcc-mkl make.inc
-echo 'LIBDIR += -L$(MKLROOT)/lib' >> make.inc
-echo 'LIB += -Wl,--enable-new-dtags -Wl,--rpath,/opt/rocm/lib -Wl,--rpath,$(MKLROOT)/lib -Wl,--rpath,/opt/rocm/magma/lib' >> make.inc
-echo 'DEVCCFLAGS += --gpu-max-threads-per-block=256' >> make.inc
-export PATH="${PATH}:/opt/rocm/bin"
-if [[ -n "$PYTORCH_ROCM_ARCH" ]]; then
-  amdgpu_targets=`echo $PYTORCH_ROCM_ARCH | sed 's/;/ /g'`
-else
-  amdgpu_targets=`rocm_agent_enumerator | grep -v gfx000 | sort -u | xargs`
-fi
-for arch in $amdgpu_targets; do
-  echo "DEVCCFLAGS += --amdgpu-target=$arch" >> make.inc
-done
-# hipcc with openmp flag may cause isnan() on __device__ not to be found; depending on context, compiler may attempt to match with host definition
-sed -i 's/^FOPENMP/#FOPENMP/g' make.inc
-make -f make.gen.hipMAGMA -j $(nproc)
-LANG=C.UTF-8 make lib/libmagma.so -j $(nproc) MKLROOT=/opt/conda
-make testing/testing_dgemm -j $(nproc) MKLROOT=/opt/conda
-popd
-mv magma /opt/rocm

.circleci/docker/common/install_ucc.sh

@@ -1,48 +0,0 @@
-#!/bin/bash
-
-set -ex
-
-if [[ -d "/usr/local/cuda/" ]];  then
-  with_cuda=/usr/local/cuda/
-else
-  with_cuda=no
-fi
-
-function install_ucx() {
-  set -ex
-  git clone --recursive https://github.com/openucx/ucx.git
-  pushd ucx
-  git checkout ${UCX_COMMIT}
-  git submodule update --init --recursive
-
-  ./autogen.sh
-  ./configure --prefix=$UCX_HOME      \
-      --enable-mt                     \
-      --with-cuda=$with_cuda          \
-      --enable-profiling              \
-      --enable-stats
-  time make -j
-  sudo make install
-
-  popd
-  rm -rf ucx
-}
-
-function install_ucc() {
-  set -ex
-  git clone --recursive https://github.com/openucx/ucc.git
-  pushd ucc
-  git checkout ${UCC_COMMIT}
-  git submodule update --init --recursive
-
-  ./autogen.sh
-  ./configure --prefix=$UCC_HOME --with-ucx=$UCX_HOME --with-cuda=$with_cuda
-  time make -j
-  sudo make install
-
-  popd
-  rm -rf ucc
-}
-
-install_ucx
-install_ucc

.circleci/docker/requirements-ci.txt

@@ -41,7 +41,7 @@ flatbuffers==2.0
 #Pinned versions:
 #test that import:
 
-hypothesis==5.35.1
+hypothesis==4.53.2
 # Pin hypothesis to avoid flakiness: https://github.com/pytorch/pytorch/issues/31136
 #Description: advanced library for generating parametrized tests
 #Pinned versions: 3.44.6, 4.53.2
@@ -80,17 +80,17 @@ librosa>=0.6.2
 #Pinned versions:
 #test that import:
 
-mypy==0.960
+mypy==0.812
 # Pin MyPy version because new errors are likely to appear with each release
 #Description: linter
-#Pinned versions: 0.960
+#Pinned versions: 0.812
 #test that import: test_typing.py, test_type_hints.py
 
-networkx==2.6.3
+#networkx
 #Description: creation, manipulation, and study of
 #the structure, dynamics, and functions of complex networks
-#Pinned versions: 2.6.3 (latest version that works with Python 3.7+)
-#test that import: functorch
+#Pinned versions: 2.0
+#test that import:
 
 #ninja
 #Description: build system.  Note that it install from
@@ -100,7 +100,6 @@ networkx==2.6.3
 
 numba==0.49.0 ; python_version < "3.9"
 numba==0.54.1 ; python_version == "3.9"
-numba==0.55.2 ; python_version == "3.10"
 #Description: Just-In-Time Compiler for Numerical Functions
 #Pinned versions: 0.54.1, 0.49.0, <=0.49.1
 #test that import: test_numba_integration.py
@@ -124,19 +123,14 @@ numba==0.55.2 ; python_version == "3.10"
 #Pinned versions: 1.9.0
 #test that import:
 
-opt-einsum==3.3
-#Description: Python library to optimize tensor contraction order, used in einsum
-#Pinned versions: 3.3
-#test that import: test_linalg.py
-
 #pillow
 #Description:  Python Imaging Library fork
 #Pinned versions:
 #test that import:
 
-protobuf==3.20.2
+#protobuf
 #Description:  Google’s data interchange format
-#Pinned versions: 3.20.1
+#Pinned versions:
 #test that import: test_tensorboard.py
 
 psutil
@@ -149,21 +143,6 @@ pytest
 #Pinned versions:
 #test that import: test_typing.py, test_cpp_extensions_aot.py, run_test.py
 
-pytest-xdist
-#Description: plugin for running pytest in parallel
-#Pinned versions:
-#test that import:
-
-pytest-shard
-#Description: plugin spliting up tests in pytest
-#Pinned versions:
-#test that import:
-
-pytest-rerunfailures
-#Description: plugin for rerunning tests in pytest
-#Pinned versions:
-#test that import:
-
 #pytest-benchmark
 #Description: fixture for benchmarking code
 #Pinned versions: 3.2.3
@@ -174,16 +153,6 @@ pytest-rerunfailures
 #Pinned versions:
 #test that import:
 
-xdoctest==1.0.2
-#Description: runs doctests in pytest
-#Pinned versions: 1.0.2
-#test that import:
-
-pygments==2.12.0
-#Description: support doctest highlighting
-#Pinned versions: 2.12.0
-#test that import: the doctests
-
 #PyYAML
 #Description: data serialization format
 #Pinned versions:
@@ -209,8 +178,7 @@ scikit-image
 #Pinned versions: 0.20.3
 #test that import:
 
-scipy==1.6.3 ; python_version < "3.10"
-scipy==1.8.1 ; python_version == "3.10"
+scipy==1.6.3
 # Pin SciPy because of failing distribution tests (see #60347)
 #Description: scientific python
 #Pinned versions: 1.6.3

.circleci/docker/ubuntu-cuda/Dockerfile

@@ -11,96 +11,80 @@ ENV DEBIAN_FRONTEND noninteractive
 
 # Install common dependencies (so that this step can be cached separately)
 ARG EC2
-COPY ./common/install_base.sh install_base.sh
+ADD ./common/install_base.sh install_base.sh
 RUN bash ./install_base.sh && rm install_base.sh
 
 # Install user
-COPY ./common/install_user.sh install_user.sh
+ADD ./common/install_user.sh install_user.sh
 RUN bash ./install_user.sh && rm install_user.sh
 
 # Install katex
 ARG KATEX
-COPY ./common/install_docs_reqs.sh install_docs_reqs.sh
-RUN bash ./install_docs_reqs.sh && rm install_docs_reqs.sh
+ADD ./common/install_katex.sh install_katex.sh
+RUN bash ./install_katex.sh && rm install_katex.sh
 
 # Install conda and other packages (e.g., numpy, pytest)
 ENV PATH /opt/conda/bin:$PATH
 ARG ANACONDA_PYTHON_VERSION
-COPY requirements-ci.txt /opt/conda/requirements-ci.txt
-COPY ./common/install_conda.sh install_conda.sh
+ADD requirements-ci.txt /opt/conda/requirements-ci.txt
+ADD ./common/install_conda.sh install_conda.sh
 RUN bash ./install_conda.sh && rm install_conda.sh
 RUN rm /opt/conda/requirements-ci.txt
 
 # Install gcc
 ARG GCC_VERSION
-COPY ./common/install_gcc.sh install_gcc.sh
+ADD ./common/install_gcc.sh install_gcc.sh
 RUN bash ./install_gcc.sh && rm install_gcc.sh
 
 # Install clang
 ARG CLANG_VERSION
-COPY ./common/install_clang.sh install_clang.sh
+ADD ./common/install_clang.sh install_clang.sh
 RUN bash ./install_clang.sh && rm install_clang.sh
 
 # (optional) Install protobuf for ONNX
 ARG PROTOBUF
-COPY ./common/install_protobuf.sh install_protobuf.sh
+ADD ./common/install_protobuf.sh install_protobuf.sh
 RUN if [ -n "${PROTOBUF}" ]; then bash ./install_protobuf.sh; fi
 RUN rm install_protobuf.sh
 ENV INSTALLED_PROTOBUF ${PROTOBUF}
 
 # (optional) Install database packages like LMDB and LevelDB
 ARG DB
-COPY ./common/install_db.sh install_db.sh
+ADD ./common/install_db.sh install_db.sh
 RUN if [ -n "${DB}" ]; then bash ./install_db.sh; fi
 RUN rm install_db.sh
 ENV INSTALLED_DB ${DB}
 
 # (optional) Install vision packages like OpenCV and ffmpeg
 ARG VISION
-COPY ./common/install_vision.sh install_vision.sh
+ADD ./common/install_vision.sh install_vision.sh
 RUN if [ -n "${VISION}" ]; then bash ./install_vision.sh; fi
 RUN rm install_vision.sh
 ENV INSTALLED_VISION ${VISION}
 
-# (optional) Install UCC
-ARG UCX_COMMIT
-ARG UCC_COMMIT
-ENV UCX_COMMIT $UCX_COMMIT
-ENV UCC_COMMIT $UCC_COMMIT
-ENV UCX_HOME /usr
-ENV UCC_HOME /usr
-ADD ./common/install_ucc.sh install_ucc.sh
-RUN if [ -n "${UCX_COMMIT}" ] && [ -n "${UCC_COMMIT}" ]; then bash ./install_ucc.sh; fi
-RUN rm install_ucc.sh
-
-COPY ./common/install_openssl.sh install_openssl.sh
+ADD ./common/install_openssl.sh install_openssl.sh
 ENV OPENSSL_ROOT_DIR /opt/openssl
 RUN bash ./install_openssl.sh
-ENV OPENSSL_DIR /opt/openssl
 
 # (optional) Install non-default CMake version
 ARG CMAKE_VERSION
-COPY ./common/install_cmake.sh install_cmake.sh
+ADD ./common/install_cmake.sh install_cmake.sh
 RUN if [ -n "${CMAKE_VERSION}" ]; then bash ./install_cmake.sh; fi
 RUN rm install_cmake.sh
 
 # Install ccache/sccache (do this last, so we get priority in PATH)
-COPY ./common/install_cache.sh install_cache.sh
+ADD ./common/install_cache.sh install_cache.sh
 ENV PATH /opt/cache/bin:$PATH
-# See https://github.com/pytorch/pytorch/issues/82174
-# TODO(sdym@fb.com):
-# check if this is needed after full off Xenial migration
-ENV CARGO_NET_GIT_FETCH_WITH_CLI true
 RUN bash ./install_cache.sh && rm install_cache.sh
 ENV CMAKE_CUDA_COMPILER_LAUNCHER=/opt/cache/bin/sccache
 
 # Add jni.h for java host build
-COPY ./common/install_jni.sh install_jni.sh
-COPY ./java/jni.h jni.h
+ADD ./common/install_jni.sh install_jni.sh
+ADD ./java/jni.h jni.h
 RUN bash ./install_jni.sh && rm install_jni.sh
 
 # Install Open MPI for CUDA
-COPY ./common/install_openmpi.sh install_openmpi.sh
+ADD ./common/install_openmpi.sh install_openmpi.sh
 RUN if [ -n "${CUDA_VERSION}" ]; then bash install_openmpi.sh; fi
 RUN rm install_openmpi.sh
 
@@ -118,14 +102,9 @@ COPY --from=pytorch/llvm:9.0.1 /opt/llvm /opt/llvm
 
 # Install CUDNN
 ARG CUDNN_VERSION
-ARG CUDA_VERSION
-COPY ./common/install_cudnn.sh install_cudnn.sh
+ADD ./common/install_cudnn.sh install_cudnn.sh
 RUN if [ "${CUDNN_VERSION}" -eq 8 ]; then bash install_cudnn.sh; fi
 RUN rm install_cudnn.sh
 
-# Delete /usr/local/cuda-11.X/cuda-11.X symlinks
-RUN if [ -h /usr/local/cuda-11.6/cuda-11.6 ]; then rm /usr/local/cuda-11.6/cuda-11.6; fi
-RUN if [ -h /usr/local/cuda-11.7/cuda-11.7 ]; then rm /usr/local/cuda-11.7/cuda-11.7; fi
-
 USER jenkins
 CMD ["bash"]

.circleci/docker/ubuntu-rocm/Dockerfile

@@ -12,61 +12,58 @@ ENV PYTORCH_ROCM_ARCH ${PYTORCH_ROCM_ARCH}
 
 # Install common dependencies (so that this step can be cached separately)
 ARG EC2
-COPY ./common/install_base.sh install_base.sh
+ADD ./common/install_base.sh install_base.sh
 RUN bash ./install_base.sh && rm install_base.sh
 
 # Install clang
 ARG LLVMDEV
 ARG CLANG_VERSION
-COPY ./common/install_clang.sh install_clang.sh
+ADD ./common/install_clang.sh install_clang.sh
 RUN bash ./install_clang.sh && rm install_clang.sh
 
 # Install user
-COPY ./common/install_user.sh install_user.sh
+ADD ./common/install_user.sh install_user.sh
 RUN bash ./install_user.sh && rm install_user.sh
 
 # Install conda and other packages (e.g., numpy, pytest)
 ENV PATH /opt/conda/bin:$PATH
 ARG ANACONDA_PYTHON_VERSION
-COPY requirements-ci.txt /opt/conda/requirements-ci.txt
-COPY ./common/install_conda.sh install_conda.sh
+ADD requirements-ci.txt /opt/conda/requirements-ci.txt
+ADD ./common/install_conda.sh install_conda.sh
 RUN bash ./install_conda.sh && rm install_conda.sh
 RUN rm /opt/conda/requirements-ci.txt
 
 # Install gcc
 ARG GCC_VERSION
-COPY ./common/install_gcc.sh install_gcc.sh
+ADD ./common/install_gcc.sh install_gcc.sh
 RUN bash ./install_gcc.sh && rm install_gcc.sh
 
 # (optional) Install protobuf for ONNX
 ARG PROTOBUF
-COPY ./common/install_protobuf.sh install_protobuf.sh
+ADD ./common/install_protobuf.sh install_protobuf.sh
 RUN if [ -n "${PROTOBUF}" ]; then bash ./install_protobuf.sh; fi
 RUN rm install_protobuf.sh
 ENV INSTALLED_PROTOBUF ${PROTOBUF}
 
 # (optional) Install database packages like LMDB and LevelDB
 ARG DB
-COPY ./common/install_db.sh install_db.sh
+ADD ./common/install_db.sh install_db.sh
 RUN if [ -n "${DB}" ]; then bash ./install_db.sh; fi
 RUN rm install_db.sh
 ENV INSTALLED_DB ${DB}
 
 # (optional) Install vision packages like OpenCV and ffmpeg
 ARG VISION
-COPY ./common/install_vision.sh install_vision.sh
+ADD ./common/install_vision.sh install_vision.sh
 RUN if [ -n "${VISION}" ]; then bash ./install_vision.sh; fi
 RUN rm install_vision.sh
 ENV INSTALLED_VISION ${VISION}
 
 # Install rocm
 ARG ROCM_VERSION
-COPY ./common/install_rocm.sh install_rocm.sh
+ADD ./common/install_rocm.sh install_rocm.sh
 RUN bash ./install_rocm.sh
 RUN rm install_rocm.sh
-COPY ./common/install_rocm_magma.sh install_rocm_magma.sh
-RUN bash ./install_rocm_magma.sh
-RUN rm install_rocm_magma.sh
 ENV PATH /opt/rocm/bin:$PATH
 ENV PATH /opt/rocm/hcc/bin:$PATH
 ENV PATH /opt/rocm/hip/bin:$PATH
@@ -78,18 +75,18 @@ ENV LC_ALL C.UTF-8
 
 # (optional) Install non-default CMake version
 ARG CMAKE_VERSION
-COPY ./common/install_cmake.sh install_cmake.sh
+ADD ./common/install_cmake.sh install_cmake.sh
 RUN if [ -n "${CMAKE_VERSION}" ]; then bash ./install_cmake.sh; fi
 RUN rm install_cmake.sh
 
 # (optional) Install non-default Ninja version
 ARG NINJA_VERSION
-COPY ./common/install_ninja.sh install_ninja.sh
+ADD ./common/install_ninja.sh install_ninja.sh
 RUN if [ -n "${NINJA_VERSION}" ]; then bash ./install_ninja.sh; fi
 RUN rm install_ninja.sh
 
 # Install ccache/sccache (do this last, so we get priority in PATH)
-COPY ./common/install_cache.sh install_cache.sh
+ADD ./common/install_cache.sh install_cache.sh
 ENV PATH /opt/cache/bin:$PATH
 RUN bash ./install_cache.sh && rm install_cache.sh
 

.circleci/docker/ubuntu/Dockerfile

@@ -6,86 +6,67 @@ ARG UBUNTU_VERSION
 
 ENV DEBIAN_FRONTEND noninteractive
 
-ARG CLANG_VERSION
-
 # Install common dependencies (so that this step can be cached separately)
 ARG EC2
-COPY ./common/install_base.sh install_base.sh
+ADD ./common/install_base.sh install_base.sh
 RUN bash ./install_base.sh && rm install_base.sh
 
 # Install clang
 ARG LLVMDEV
-COPY ./common/install_clang.sh install_clang.sh
+ARG CLANG_VERSION
+ADD ./common/install_clang.sh install_clang.sh
 RUN bash ./install_clang.sh && rm install_clang.sh
 
 # (optional) Install thrift.
 ARG THRIFT
-COPY ./common/install_thrift.sh install_thrift.sh
+ADD ./common/install_thrift.sh install_thrift.sh
 RUN if [ -n "${THRIFT}" ]; then bash ./install_thrift.sh; fi
 RUN rm install_thrift.sh
 ENV INSTALLED_THRIFT ${THRIFT}
 
 # Install user
-COPY ./common/install_user.sh install_user.sh
+ADD ./common/install_user.sh install_user.sh
 RUN bash ./install_user.sh && rm install_user.sh
 
 # Install katex
 ARG KATEX
-COPY ./common/install_docs_reqs.sh install_docs_reqs.sh
-RUN bash ./install_docs_reqs.sh && rm install_docs_reqs.sh
+ADD ./common/install_katex.sh install_katex.sh
+RUN bash ./install_katex.sh && rm install_katex.sh
 
 # Install conda and other packages (e.g., numpy, pytest)
 ENV PATH /opt/conda/bin:$PATH
 ARG ANACONDA_PYTHON_VERSION
-COPY requirements-ci.txt /opt/conda/requirements-ci.txt
-COPY ./common/install_conda.sh install_conda.sh
+ADD requirements-ci.txt /opt/conda/requirements-ci.txt
+ADD ./common/install_conda.sh install_conda.sh
 RUN bash ./install_conda.sh && rm install_conda.sh
 RUN rm /opt/conda/requirements-ci.txt
 
 # Install gcc
 ARG GCC_VERSION
-COPY ./common/install_gcc.sh install_gcc.sh
+ADD ./common/install_gcc.sh install_gcc.sh
 RUN bash ./install_gcc.sh && rm install_gcc.sh
 
 # Install lcov for C++ code coverage
-COPY ./common/install_lcov.sh install_lcov.sh
+ADD ./common/install_lcov.sh install_lcov.sh
 RUN  bash ./install_lcov.sh && rm install_lcov.sh
 
-# Install cuda and cudnn
-ARG CUDA_VERSION
-RUN wget -q https://raw.githubusercontent.com/pytorch/builder/main/common/install_cuda.sh -O install_cuda.sh
-RUN bash ./install_cuda.sh ${CUDA_VERSION} && rm install_cuda.sh
-ENV DESIRED_CUDA ${CUDA_VERSION}
-ENV PATH /usr/local/nvidia/bin:/usr/local/cuda/bin:$PATH
-
-# (optional) Install UCC
-ARG UCX_COMMIT
-ARG UCC_COMMIT
-ENV UCX_COMMIT $UCX_COMMIT
-ENV UCC_COMMIT $UCC_COMMIT
-ENV UCX_HOME /usr
-ENV UCC_HOME /usr
-ADD ./common/install_ucc.sh install_ucc.sh
-RUN if [ -n "${UCX_COMMIT}" ] && [ -n "${UCC_COMMIT}" ]; then bash ./install_ucc.sh; fi
-RUN rm install_ucc.sh
-
 # (optional) Install protobuf for ONNX
 ARG PROTOBUF
-COPY ./common/install_protobuf.sh install_protobuf.sh
+ADD ./common/install_protobuf.sh install_protobuf.sh
 RUN if [ -n "${PROTOBUF}" ]; then bash ./install_protobuf.sh; fi
 RUN rm install_protobuf.sh
 ENV INSTALLED_PROTOBUF ${PROTOBUF}
 
 # (optional) Install database packages like LMDB and LevelDB
 ARG DB
-COPY ./common/install_db.sh install_db.sh
+ADD ./common/install_db.sh install_db.sh
 RUN if [ -n "${DB}" ]; then bash ./install_db.sh; fi
 RUN rm install_db.sh
 ENV INSTALLED_DB ${DB}
 
 # (optional) Install vision packages like OpenCV and ffmpeg
 ARG VISION
-COPY ./common/install_vision.sh install_vision.sh
+ADD ./common/install_vision.sh install_vision.sh
 RUN if [ -n "${VISION}" ]; then bash ./install_vision.sh; fi
 RUN rm install_vision.sh
 ENV INSTALLED_VISION ${VISION}
@@ -94,9 +75,9 @@ ENV INSTALLED_VISION ${VISION}
 ARG ANDROID
 ARG ANDROID_NDK
 ARG GRADLE_VERSION
-COPY ./common/install_android.sh install_android.sh
-COPY ./android/AndroidManifest.xml AndroidManifest.xml
-COPY ./android/build.gradle build.gradle
+ADD ./common/install_android.sh install_android.sh
+ADD ./android/AndroidManifest.xml AndroidManifest.xml
+ADD ./android/build.gradle build.gradle
 RUN if [ -n "${ANDROID}" ]; then bash ./install_android.sh; fi
 RUN rm install_android.sh
 RUN rm AndroidManifest.xml
@@ -105,53 +86,42 @@ ENV INSTALLED_ANDROID ${ANDROID}
 
 # (optional) Install Vulkan SDK
 ARG VULKAN_SDK_VERSION
-COPY ./common/install_vulkan_sdk.sh install_vulkan_sdk.sh
+ADD ./common/install_vulkan_sdk.sh install_vulkan_sdk.sh
 RUN if [ -n "${VULKAN_SDK_VERSION}" ]; then bash ./install_vulkan_sdk.sh; fi
 RUN rm install_vulkan_sdk.sh
 
 # (optional) Install swiftshader
 ARG SWIFTSHADER
-COPY ./common/install_swiftshader.sh install_swiftshader.sh
+ADD ./common/install_swiftshader.sh install_swiftshader.sh
 RUN if [ -n "${SWIFTSHADER}" ]; then bash ./install_swiftshader.sh; fi
 RUN rm install_swiftshader.sh
 
 # (optional) Install non-default CMake version
 ARG CMAKE_VERSION
-COPY ./common/install_cmake.sh install_cmake.sh
+ADD ./common/install_cmake.sh install_cmake.sh
 RUN if [ -n "${CMAKE_VERSION}" ]; then bash ./install_cmake.sh; fi
 RUN rm install_cmake.sh
 
 # (optional) Install non-default Ninja version
 ARG NINJA_VERSION
-COPY ./common/install_ninja.sh install_ninja.sh
+ADD ./common/install_ninja.sh install_ninja.sh
 RUN if [ -n "${NINJA_VERSION}" ]; then bash ./install_ninja.sh; fi
 RUN rm install_ninja.sh
 
-COPY ./common/install_openssl.sh install_openssl.sh
+ADD ./common/install_openssl.sh install_openssl.sh
 RUN bash ./install_openssl.sh
 ENV OPENSSL_ROOT_DIR /opt/openssl
-ENV OPENSSL_DIR /opt/openssl
-RUN rm install_openssl.sh
 
 # Install ccache/sccache (do this last, so we get priority in PATH)
-COPY ./common/install_cache.sh install_cache.sh
+ADD ./common/install_cache.sh install_cache.sh
 ENV PATH /opt/cache/bin:$PATH
-# See https://github.com/pytorch/pytorch/issues/82174
-# TODO(sdym@fb.com):
-# check if this is needed after full off Xenial migration
-ENV CARGO_NET_GIT_FETCH_WITH_CLI true
 RUN bash ./install_cache.sh && rm install_cache.sh
 
 # Add jni.h for java host build
-COPY ./common/install_jni.sh install_jni.sh
-COPY ./java/jni.h jni.h
+ADD ./common/install_jni.sh install_jni.sh
+ADD ./java/jni.h jni.h
 RUN bash ./install_jni.sh && rm install_jni.sh
 
-# Install Open MPI for CUDA
-COPY ./common/install_openmpi.sh install_openmpi.sh
-RUN if [ -n "${CUDA_VERSION}" ]; then bash install_openmpi.sh; fi
-RUN rm install_openmpi.sh
-
 # Include BUILD_ENVIRONMENT environment variable in image
 ARG BUILD_ENVIRONMENT
 ENV BUILD_ENVIRONMENT ${BUILD_ENVIRONMENT}
@@ -159,10 +129,5 @@ ENV BUILD_ENVIRONMENT ${BUILD_ENVIRONMENT}
 # Install LLVM dev version (Defined in the pytorch/builder github repository)
 COPY --from=pytorch/llvm:9.0.1 /opt/llvm /opt/llvm
 
-# AWS specific CUDA build guidance
-ENV TORCH_CUDA_ARCH_LIST Maxwell
-ENV TORCH_NVCC_FLAGS "-Xfatbin -compress-all"
-ENV CUDA_PATH /usr/local/cuda
-
 USER jenkins
 CMD ["bash"]

.circleci/generate_config_yml.py

@@ -14,9 +14,6 @@ import cimodel.data.simple.docker_definitions
 import cimodel.data.simple.mobile_definitions
 import cimodel.data.simple.nightly_ios
 import cimodel.data.simple.anaconda_prune_defintions
-import cimodel.data.simple.macos_definitions
-import cimodel.data.simple.upload_test_stats_definition
-import cimodel.data.simple.ios_definitions
 import cimodel.lib.miniutils as miniutils
 import cimodel.lib.miniyaml as miniyaml
 
@@ -73,7 +70,6 @@ class Header(object):
         for line in filter(None, lines):
             output_filehandle.write(line + "\n")
 
-
 def _for_all_items(items, functor) -> None:
     if isinstance(items, list):
         for item in items:
@@ -82,7 +78,6 @@ def _for_all_items(items, functor) -> None:
         item_type, item = next(iter(items.items()))
         functor(item_type, item)
 
-
 def filter_master_only_jobs(items):
     def _is_main_or_master_item(item):
         filters = item.get('filters', None)
@@ -121,7 +116,6 @@ def filter_master_only_jobs(items):
     _for_all_items(items, _save_requires_if_master)
     return _do_filtering(items)
 
-
 def generate_required_docker_images(items):
     required_docker_images = set()
 
@@ -137,15 +131,11 @@ def generate_required_docker_images(items):
     _for_all_items(items, _requires_docker_image)
     return required_docker_images
 
-
 def gen_build_workflows_tree():
     build_workflows_functions = [
         cimodel.data.simple.mobile_definitions.get_workflow_jobs,
         cimodel.data.simple.nightly_ios.get_workflow_jobs,
         cimodel.data.simple.anaconda_prune_defintions.get_workflow_jobs,
-        cimodel.data.simple.macos_definitions.get_new_workflow_jobs,
-        cimodel.data.simple.upload_test_stats_definition.get_workflow_job,
-        cimodel.data.simple.ios_definitions.get_workflow_jobs,
     ]
     build_jobs = [f() for f in build_workflows_functions]
     build_jobs.extend(

.circleci/scripts/binary_checkout.sh

@@ -62,7 +62,7 @@ git --no-pager log --max-count 1
 popd
 
 # Clone the Builder master repo
-retry git clone -q https://github.com/pytorch/builder.git -b release/1.13 "$BUILDER_ROOT"
+retry git clone -q https://github.com/pytorch/builder.git -b release/1.12 "$BUILDER_ROOT"
 pushd "$BUILDER_ROOT"
 echo "Using builder from "
 git --no-pager log --max-count 1

.circleci/scripts/binary_ios_test.sh

@@ -1,19 +1,30 @@
 #!/bin/bash
 set -ex -o pipefail
 
-if ! [ "$IOS_PLATFORM" == "SIMULATOR" ]; then
-    exit 0
-fi
-
 echo ""
 echo "DIR: $(pwd)"
 PROJ_ROOT=/Users/distiller/project
 cd ${PROJ_ROOT}/ios/TestApp
 # install fastlane
 sudo gem install bundler && bundle install
+# install certificates
+echo "${IOS_CERT_KEY_2022}" >> cert.txt
+base64 --decode cert.txt -o Certificates.p12
+rm cert.txt
+bundle exec fastlane install_root_cert
+bundle exec fastlane install_dev_cert
+# install the provisioning profile
+PROFILE=PyTorch_CI_2022.mobileprovision
+PROVISIONING_PROFILES=~/Library/MobileDevice/Provisioning\ Profiles
+mkdir -pv "${PROVISIONING_PROFILES}"
+cd "${PROVISIONING_PROFILES}"
+echo "${IOS_SIGN_KEY_2022}" >> cert.txt
+base64 --decode cert.txt -o ${PROFILE}
+rm cert.txt
 # run the ruby build script
 if ! [ -x "$(command -v xcodebuild)" ]; then
     echo 'Error: xcodebuild is not installed.'
     exit 1
 fi
-ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM}
+PROFILE=PyTorch_CI_2022
+ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM} -c ${PROFILE} -t ${IOS_DEV_TEAM_ID}

.circleci/scripts/binary_ios_upload.sh

@@ -33,7 +33,7 @@ fi
 cp ${PROJ_ROOT}/LICENSE ${ZIP_DIR}/
 # zip the library
 export DATE="$(date -u +%Y%m%d)"
-export IOS_NIGHTLY_BUILD_VERSION="1.13.0.${DATE}"
+export IOS_NIGHTLY_BUILD_VERSION="1.12.0.${DATE}"
 if [ "${BUILD_LITE_INTERPRETER}" == "1" ]; then
     # libtorch_lite_ios_nightly_1.11.0.20210810.zip
     ZIPFILE="libtorch_lite_ios_nightly_${IOS_NIGHTLY_BUILD_VERSION}.zip"

.circleci/scripts/binary_linux_test.sh

@@ -53,7 +53,9 @@ if [[ "\$python_nodot" = *39*  ]]; then
   NUMPY_PIN=">=1.20"
 fi
 
-
+if [[ "$DESIRED_CUDA" == "cu116" ]]; then
+  EXTRA_CONDA_FLAGS="-c=conda-forge"
+fi
 
 # Move debug wheels out of the the package dir so they don't get installed
 mkdir -p /tmp/debug_final_pkgs
@@ -86,14 +88,13 @@ if [[ "$PACKAGE_TYPE" == conda ]]; then
     if [[ "$DESIRED_CUDA" == 'cpu' ]]; then
       retry conda install -c pytorch -y cpuonly
     else
-
-      cu_ver="${DESIRED_CUDA:2:2}.${DESIRED_CUDA:4}"
-      CUDA_PACKAGE="cudatoolkit"
-      if [[ "$DESIRED_CUDA" == "cu116" || "$DESIRED_CUDA" == "cu117" ]]; then
-        CUDA_PACKAGE="cuda"
+      # DESIRED_CUDA is in format cu90 or cu102
+      if [[ "${#DESIRED_CUDA}" == 4 ]]; then
+        cu_ver="${DESIRED_CUDA:2:1}.${DESIRED_CUDA:3}"
+      else
+        cu_ver="${DESIRED_CUDA:2:2}.${DESIRED_CUDA:4}"
       fi
-
-      retry conda install \${EXTRA_CONDA_FLAGS} -yq -c nvidia -c pytorch "\${CUDA_PACKAGE}=\${cu_ver}"
+      retry conda install \${EXTRA_CONDA_FLAGS} -yq -c nvidia -c pytorch "cudatoolkit=\${cu_ver}"
     fi
     conda install \${EXTRA_CONDA_FLAGS} -y "\$pkg" --offline
   )

.circleci/scripts/binary_macos_build.sh

@@ -4,7 +4,7 @@ set -eux -o pipefail
 source "${BINARY_ENV_FILE:-/Users/distiller/project/env}"
 mkdir -p "$PYTORCH_FINAL_PACKAGE_DIR"
 
-if [[ -z "${GITHUB_ACTIONS:-}" ]]; then
+if [[ -z "${IS_GHA:-}" ]]; then
   export PATH="${workdir:-${HOME}}/miniconda/bin:${PATH}"
 fi
 

.circleci/scripts/binary_populate_env.sh

@@ -5,7 +5,7 @@ export TZ=UTC
 tagged_version() {
   # Grabs version from either the env variable CIRCLE_TAG
   # or the pytorch git described version
-  if [[ "$OSTYPE" == "msys" &&  -z "${GITHUB_ACTIONS:-}" ]]; then
+  if [[ "$OSTYPE" == "msys" &&  -z "${IS_GHA:-}" ]]; then
     GIT_DIR="${workdir}/p/.git"
   else
     GIT_DIR="${workdir}/pytorch/.git"
@@ -23,12 +23,50 @@ tagged_version() {
   fi
 }
 
-envfile=${BINARY_ENV_FILE:-/tmp/env}
-if [[ -n "${PYTORCH_ROOT}"  ]]; then
-  workdir=$(dirname "${PYTORCH_ROOT}")
+# These are only relevant for CircleCI
+# TODO: Remove these later once migrated fully to GHA
+if [[ -z ${IS_GHA:-} ]]; then
+  # We need to write an envfile to persist these variables to following
+  # steps, but the location of the envfile depends on the circleci executor
+  if [[ "$(uname)" == Darwin ]]; then
+    # macos executor (builds and tests)
+    workdir="/Users/distiller/project"
+  elif [[ "$OSTYPE" == "msys" ]]; then
+    # windows executor (builds and tests)
+    workdir="/c/w"
+  elif [[ -d "/home/circleci/project" ]]; then
+    # machine executor (binary tests)
+    workdir="/home/circleci/project"
+  else
+    # docker executor (binary builds)
+    workdir="/"
+  fi
+  envfile="$workdir/env"
+  touch "$envfile"
+  chmod +x "$envfile"
+
+  # Parse the BUILD_ENVIRONMENT to package type, python, and cuda
+  configs=($BUILD_ENVIRONMENT)
+  export PACKAGE_TYPE="${configs[0]}"
+  export DESIRED_PYTHON="${configs[1]}"
+  export DESIRED_CUDA="${configs[2]}"
+  if [[ "${OSTYPE}" == "msys" ]]; then
+    export DESIRED_DEVTOOLSET=""
+    export LIBTORCH_CONFIG="${configs[3]:-}"
+    if [[ "$LIBTORCH_CONFIG" == 'debug' ]]; then
+      export DEBUG=1
+    fi
+  else
+    export DESIRED_DEVTOOLSET="${configs[3]:-}"
+  fi
 else
-  # docker executor (binary builds)
-  workdir="/"
+  envfile=${BINARY_ENV_FILE:-/tmp/env}
+  if [[ -n "${PYTORCH_ROOT}"  ]]; then
+    workdir=$(dirname "${PYTORCH_ROOT}")
+  else
+    # docker executor (binary builds)
+    workdir="/"
+  fi
 fi
 
 if [[ "$PACKAGE_TYPE" == 'libtorch' ]]; then
@@ -59,7 +97,7 @@ PIP_UPLOAD_FOLDER='nightly/'
 # We put this here so that OVERRIDE_PACKAGE_VERSION below can read from it
 export DATE="$(date -u +%Y%m%d)"
 #TODO: We should be pulling semver version from the base version.txt
-BASE_BUILD_VERSION="1.13.0.dev$DATE"
+BASE_BUILD_VERSION="1.12.0.dev$DATE"
 # Change BASE_BUILD_VERSION to git tag when on a git tag
 # Use 'git -C' to make doubly sure we're in the correct directory for checking
 # the git tag
@@ -76,11 +114,6 @@ if [[ "$(uname)" == 'Darwin' ]] || [[ "$PACKAGE_TYPE" == conda ]]; then
 else
   export PYTORCH_BUILD_VERSION="${BASE_BUILD_VERSION}+$DESIRED_CUDA"
 fi
-
-if [[ -n "${PYTORCH_EXTRA_INSTALL_REQUIREMENTS:-}" ]]; then
-  export PYTORCH_BUILD_VERSION="${PYTORCH_BUILD_VERSION}-with-pypi-cudnn"
-fi
-
 export PYTORCH_BUILD_NUMBER=1
 
 
@@ -129,9 +162,9 @@ if [[ "${OSTYPE}" == "msys" ]]; then
 else
   export DESIRED_DEVTOOLSET="${DESIRED_DEVTOOLSET:-}"
 fi
-export PYTORCH_EXTRA_INSTALL_REQUIREMENTS="${PYTORCH_EXTRA_INSTALL_REQUIREMENTS:-}"
+
 export DATE="$DATE"
-export NIGHTLIES_DATE_PREAMBLE=1.13.0.dev
+export NIGHTLIES_DATE_PREAMBLE=1.12.0.dev
 export PYTORCH_BUILD_VERSION="$PYTORCH_BUILD_VERSION"
 export PYTORCH_BUILD_NUMBER="$PYTORCH_BUILD_NUMBER"
 export OVERRIDE_PACKAGE_VERSION="$PYTORCH_BUILD_VERSION"
@@ -167,7 +200,7 @@ if [[ "$(uname)" != Darwin ]]; then
 EOL
 fi
 
-if [[ -z "${GITHUB_ACTIONS:-}" ]]; then
+if [[ -z "${IS_GHA:-}" ]]; then
   cat >>"$envfile" <<EOL
   export workdir="$workdir"
   export MAC_PACKAGE_WORK_DIR="$workdir"

.circleci/scripts/binary_upload.sh

@@ -14,12 +14,6 @@ UPLOAD_CHANNEL=${UPLOAD_CHANNEL:-nightly}
 UPLOAD_SUBFOLDER=${UPLOAD_SUBFOLDER:-cpu}
 UPLOAD_BUCKET="s3://pytorch"
 BACKUP_BUCKET="s3://pytorch-backup"
-BUILD_NAME=${BUILD_NAME:-}
-
-# this is temporary change to upload pypi-cudnn builds to separate folder
-if [[ ${BUILD_NAME} == *with-pypi-cudnn* ]]; then
-  UPLOAD_SUBFOLDER="${UPLOAD_SUBFOLDER}_pypi_cudnn"
-fi
 
 DRY_RUN=${DRY_RUN:-enabled}
 # Don't actually do work unless explicit
@@ -30,11 +24,6 @@ if [[ "${DRY_RUN}" = "disabled" ]]; then
   AWS_S3_CP="aws s3 cp"
 fi
 
-# Sleep 2 minutes between retries for conda upload
-retry () {
-  "$@"  || (sleep 5m && "$@") || (sleep 5m && "$@") || (sleep 5m && "$@") || (sleep 5m && "$@")
-}
-
 do_backup() {
   local backup_dir
   backup_dir=$1
@@ -48,14 +37,13 @@ do_backup() {
 conda_upload() {
   (
     set -x
-    retry \
     ${ANACONDA} \
-    upload  \
-    ${PKG_DIR}/*.tar.bz2 \
-    -u "pytorch-${UPLOAD_CHANNEL}" \
-    --label main \
-    --no-progress \
-    --force
+      upload  \
+      ${PKG_DIR}/*.tar.bz2 \
+      -u "pytorch-${UPLOAD_CHANNEL}" \
+      --label main \
+      --no-progress \
+      --force
   )
 }
 

.circleci/scripts/binary_windows_build.sh

@@ -6,7 +6,7 @@ mkdir -p "$PYTORCH_FINAL_PACKAGE_DIR"
 
 export CUDA_VERSION="${DESIRED_CUDA/cu/}"
 export USE_SCCACHE=1
-export SCCACHE_BUCKET=ossci-compiler-cache
+export SCCACHE_BUCKET=ossci-compiler-cache-windows
 export SCCACHE_IGNORE_SERVER_IO_ERROR=1
 export VC_YEAR=2019
 

.circleci/scripts/build_android_gradle.sh

@@ -78,7 +78,7 @@ if [[ "${BUILD_ENVIRONMENT}" == *-gradle-build-only-x86_32* ]]; then
     GRADLE_PARAMS+=" -PABI_FILTERS=x86"
 fi
 
-if [ -n "${GRADLE_OFFLINE:-}" ]; then
+if [ -n "{GRADLE_OFFLINE:-}" ]; then
     GRADLE_PARAMS+=" --offline"
 fi
 

.circleci/scripts/cpp_doc_push_script.sh

@@ -51,6 +51,8 @@ git clone https://github.com/pytorch/cppdocs
 
 set -ex
 
+sudo apt-get -y install doxygen
+
 # Generate ATen files
 pushd "${pt_checkout}"
 pip install -r requirements.txt
@@ -98,9 +100,6 @@ git commit -m "Generate C++ docs from pytorch/pytorch@${GITHUB_SHA}" || true
 git status
 
 if [[ "${WITH_PUSH:-}" == true ]]; then
-  # push to a temp branch first to trigger CLA check and satisfy branch protections
-  git push -u origin HEAD:pytorchbot/temp-branch-cpp -f
-  sleep 30
   git push -u origin
 fi
 

.circleci/scripts/functorch_doc_push_script.sh

@@ -1,47 +0,0 @@
-#!/bin/bash
-# =================== The following code **should** be executed inside Docker container ===================
-
-# Install dependencies
-sudo apt-get -y update
-sudo apt-get -y install expect-dev
-
-# This is where the local pytorch install in the docker image is located
-pt_checkout="/var/lib/jenkins/workspace"
-source "$pt_checkout/.jenkins/pytorch/common_utils.sh"
-echo "functorch_doc_push_script.sh: Invoked with $*"
-
-set -ex
-
-version=${DOCS_VERSION:-nightly}
-echo "version: $version"
-
-# Build functorch docs
-pushd $pt_checkout/functorch/docs
-pip -q install -r requirements.txt
-make html
-popd
-
-git clone https://github.com/pytorch/functorch -b gh-pages --depth 1 functorch_ghpages
-pushd functorch_ghpages
-
-if [ $version == "master" ]; then
-  version=nightly
-fi
-
-git rm -rf "$version" || true
-mv "$pt_checkout/functorch/docs/build/html" "$version"
-
-git add "$version" || true
-git status
-git config user.email "soumith+bot@pytorch.org"
-git config user.name "pytorchbot"
-# If there aren't changes, don't make a commit; push is no-op
-git commit -m "Generate Python docs from pytorch/pytorch@${GITHUB_SHA}" || true
-git status
-
-if [[ "${WITH_PUSH:-}" == true ]]; then
-  git push -u origin gh-pages
-fi
-
-popd
-# =================== The above code **should** be executed inside Docker container ===================

.circleci/scripts/python_doc_push_script.sh

@@ -135,9 +135,6 @@ git commit -m "Generate Python docs from pytorch/pytorch@${GITHUB_SHA}" || true
 git status
 
 if [[ "${WITH_PUSH:-}" == true ]]; then
-  # push to a temp branch first to trigger CLA check and satisfy branch protections
-  git push -u origin HEAD:pytorchbot/temp-branch-py -f
-  sleep 30
   git push -u origin "${branch}"
 fi
 

.circleci/scripts/setup_ci_environment.sh

@@ -32,7 +32,7 @@ if ! command -v aws >/dev/null; then
 fi
 
 if [ -n "${USE_CUDA_DOCKER_RUNTIME:-}" ]; then
-  DRIVER_FN="NVIDIA-Linux-x86_64-515.57.run"
+  DRIVER_FN="NVIDIA-Linux-x86_64-510.60.02.run"
   wget "https://s3.amazonaws.com/ossci-linux/nvidia_driver/$DRIVER_FN"
   sudo /bin/bash "$DRIVER_FN" -s --no-drm || (sudo cat /var/log/nvidia-installer.log && false)
   nvidia-smi
@@ -66,6 +66,7 @@ add_to_env_file() {
   esac
 }
 
+add_to_env_file IN_CI 1
 add_to_env_file CI_MASTER "${CI_MASTER:-}"
 add_to_env_file COMMIT_SOURCE "${CIRCLE_BRANCH:-}"
 add_to_env_file BUILD_ENVIRONMENT "${BUILD_ENVIRONMENT}"

.circleci/scripts/windows_cuda_install.sh

@@ -14,11 +14,6 @@ case ${CUDA_VERSION} in
         cuda_installer_name="cuda_11.6.0_511.23_windows"
         cuda_install_packages="thrust_11.6 nvcc_11.6 cuobjdump_11.6 nvprune_11.6 nvprof_11.6 cupti_11.6 cublas_11.6 cublas_dev_11.6 cudart_11.6 cufft_11.6 cufft_dev_11.6 curand_11.6 curand_dev_11.6 cusolver_11.6 cusolver_dev_11.6 cusparse_11.6 cusparse_dev_11.6 npp_11.6 npp_dev_11.6 nvrtc_11.6 nvrtc_dev_11.6 nvml_dev_11.6"
         ;;
-    11.7)
-        cuda_installer_name="cuda_11.7.0_516.01_windows"
-        cuda_install_packages="thrust_11.7 nvcc_11.7 cuobjdump_11.7 nvprune_11.7 nvprof_11.7 cupti_11.7 cublas_11.7 cublas_dev_11.7 cudart_11.7 cufft_11.7 cufft_dev_11.7 curand_11.7 curand_dev_11.7 cusolver_11.7 cusolver_dev_11.7 cusparse_11.7 cusparse_dev_11.7 npp_11.7 npp_dev_11.7 nvrtc_11.7 nvrtc_dev_11.7 nvml_dev_11.7"
-        ;;
-
     *)
         echo "CUDA_VERSION $CUDA_VERSION is not supported yet"
         exit 1

.circleci/scripts/windows_cudnn_install.sh

@@ -16,10 +16,6 @@ case ${CUDA_VERSION} in
         # Use cudnn8.3 with hard-coded cuda11.5 version
         cudnn_file_name="cudnn-windows-x86_64-8.3.2.44_cuda11.5-archive"
         ;;
-    11.7)
-        # Use cudnn8.3 with hard-coded cuda11.5 version
-        cudnn_file_name="cudnn-windows-x86_64-8.5.0.96_cuda11-archive"
-        ;;
     *)
         echo "CUDA_VERSION: ${CUDA_VERSION} not supported yet"
         exit 1

.circleci/verbatim-sources/commands.yml

@@ -132,3 +132,43 @@ commands:
             else
               echo "This is not a pull request, skipping..."
             fi
+
+  upload_binary_size_for_android_build:
+    description: "Upload binary size data for Android build"
+    parameters:
+      build_type:
+        type: string
+        default: ""
+      artifacts:
+        type: string
+        default: ""
+    steps:
+      - run:
+          name: "Binary Size - Install Dependencies"
+          no_output_timeout: "5m"
+          command: |
+            retry () {
+              $* || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
+            }
+            retry pip3 install requests
+      - run:
+          name: "Binary Size - Untar Artifacts"
+          no_output_timeout: "5m"
+          command: |
+            # The artifact file is created inside docker container, which contains the result binaries.
+            # Now unpackage it into the project folder. The subsequent script will scan project folder
+            # to locate result binaries and report their sizes.
+            # If artifact file is not provided it assumes that the project folder has been mounted in
+            # the docker during build and already contains the result binaries, so this step can be skipped.
+            export ARTIFACTS="<< parameters.artifacts >>"
+            if [ -n "${ARTIFACTS}" ]; then
+              tar xf "${ARTIFACTS}" -C ~/project
+            fi
+      - run:
+          name: "Binary Size - Upload << parameters.build_type >>"
+          no_output_timeout: "5m"
+          command: |
+            cd ~/project
+            export ANDROID_BUILD_TYPE="<< parameters.build_type >>"
+            export COMMIT_TIME=$(git log --max-count=1 --format=%ct || echo 0)
+            python3 -m tools.stats.upload_binary_size_to_scuba android

.circleci/verbatim-sources/job-specs/binary-job-specs.yml

@@ -1,4 +1,243 @@
 jobs:
+  binary_linux_build:
+    <<: *binary_linux_build_params
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Build
+        no_output_timeout: "1h"
+        command: |
+            source "/pytorch/.circleci/scripts/binary_linux_build.sh"
+            # Preserve build log
+            if [ -f /pytorch/build/.ninja_log ]; then
+              cp /pytorch/build/.ninja_log /final_pkgs
+            fi
+    - run:
+        name: Output binary sizes
+        no_output_timeout: "1m"
+        command: |
+            ls -lah /final_pkgs
+    - run:
+        name: upload build & binary data
+        no_output_timeout: "5m"
+        command: |
+            source /env
+            cd /pytorch && export COMMIT_TIME=$(git log --max-count=1 --format=%ct || echo 0)
+            python3 -mpip install requests && \
+            SCRIBE_GRAPHQL_ACCESS_TOKEN=${SCRIBE_GRAPHQL_ACCESS_TOKEN} \
+            python3 -m tools.stats.upload_binary_size_to_scuba || exit 0
+    - persist_to_workspace:
+        root: /
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /final_pkgs
+
+    # This should really just be another step of the binary_linux_build job above.
+    # This isn't possible right now b/c the build job uses the docker executor
+    # (otherwise they'd be really really slow) but this one uses the macine
+    # executor (b/c we have to run the docker with --runtime=nvidia and we can't do
+    # that on the docker executor)
+  binary_linux_test:
+    <<: *binary_linux_test_upload_params
+    machine:
+        image: ubuntu-2004:202104-01
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - attach_workspace:
+        at: /home/circleci/project
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Prepare test code
+        no_output_timeout: "1h"
+        command: .circleci/scripts/binary_linux_test.sh
+    - run:
+        <<: *binary_run_in_docker
+
+  binary_upload:
+    parameters:
+      package_type:
+        type: string
+        description: "What type of package we are uploading (eg. wheel, libtorch, conda)"
+        default: "wheel"
+      upload_subfolder:
+        type: string
+        description: "What subfolder to put our package into (eg. cpu, cudaX.Y, etc.)"
+        default: "cpu"
+    docker:
+      - image: continuumio/miniconda3
+    environment:
+      - DRY_RUN: disabled
+      - PACKAGE_TYPE: "<< parameters.package_type >>"
+      - UPLOAD_SUBFOLDER: "<< parameters.upload_subfolder >>"
+    steps:
+      - attach_workspace:
+          at: /tmp/workspace
+      - checkout
+      - designate_upload_channel
+      - run:
+          name: Install dependencies
+          no_output_timeout: "1h"
+          command: |
+            conda install -yq anaconda-client
+            pip install -q awscli
+      - run:
+          name: Do upload
+          no_output_timeout: "1h"
+          command: |
+            AWS_ACCESS_KEY_ID="${PYTORCH_BINARY_AWS_ACCESS_KEY_ID}" \
+              AWS_SECRET_ACCESS_KEY="${PYTORCH_BINARY_AWS_SECRET_ACCESS_KEY}" \
+              ANACONDA_API_TOKEN="${CONDA_PYTORCHBOT_TOKEN}" \
+              .circleci/scripts/binary_upload.sh
+
+  # Nighlty build smoke tests defaults
+  # These are the second-round smoke tests. These make sure that the binaries are
+  # correct from a user perspective, testing that they exist from the cloud are
+  # are runnable. Note that the pytorch repo is never cloned into these jobs
+  ##############################################################################
+  smoke_linux_test:
+    <<: *binary_linux_test_upload_params
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          cat >/home/circleci/project/ci_test_script.sh \<<EOL
+          # The following code will be executed inside Docker container
+          set -eux -o pipefail
+          /builder/smoke_test.sh
+          # The above code will be executed inside Docker container
+          EOL
+    - run:
+        <<: *binary_run_in_docker
+
+  smoke_mac_test:
+    <<: *binary_linux_test_upload_params
+    macos:
+      xcode: "12.0"
+    steps:
+      - checkout
+      - run:
+          <<: *binary_checkout
+      - run:
+          <<: *binary_populate_env
+      - brew_update
+      - run:
+          <<: *binary_install_miniconda
+      - run:
+          name: Build
+          no_output_timeout: "1h"
+          command: |
+            set -ex
+            source "/Users/distiller/project/env"
+            export "PATH=$workdir/miniconda/bin:$PATH"
+            # TODO unbuffer and ts this, but it breaks cause miniconda overwrites
+            # tclsh. But unbuffer and ts aren't that important so they're just
+            # disabled for now
+            ./builder/smoke_test.sh
+
+  binary_mac_build:
+    <<: *binary_mac_params
+    macos:
+      xcode: "12.0"
+      resource_class: "large"
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - brew_update
+    - run:
+        <<: *binary_install_miniconda
+
+    - run:
+        name: Build
+        no_output_timeout: "90m"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_build.sh"
+          cat "$script"
+          source "$script"
+
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_test.sh"
+          cat "$script"
+          source "$script"
+
+    - persist_to_workspace:
+        root: /Users/distiller/project
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /Users/distiller/project/final_pkgs
+
+  binary_macos_arm64_build:
+    <<: *binary_mac_params
+    macos:
+      xcode: "12.3.0"
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - brew_update
+    - run:
+        <<: *binary_install_miniconda
+
+    - run:
+        name: Build
+        no_output_timeout: "90m"
+        command: |
+          # Do not set -u here; there is some problem with CircleCI
+          # variable expansion with PROMPT_COMMAND
+          set -ex -o pipefail
+          export CROSS_COMPILE_ARM64=1
+          script="/Users/distiller/project/pytorch/.circleci/scripts/binary_macos_build.sh"
+          cat "$script"
+          source "$script"
+
+    - persist_to_workspace:
+        root: /Users/distiller/project
+        paths: final_pkgs
+
+    - store_artifacts:
+        path: /Users/distiller/project/final_pkgs
+
+
   binary_ios_build:
     <<: *pytorch_ios_params
     macos:
@@ -43,6 +282,90 @@ jobs:
           cat "$script"
           source "$script"
 
+  binary_windows_build:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-xlarge-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    # See Note [Workspace for CircleCI scripts] in job-specs-setup.yml
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Build
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          script="/c/w/p/.circleci/scripts/binary_windows_build.sh"
+          cat "$script"
+          source "$script"
+    - persist_to_workspace:
+        root: "C:/w"
+        paths: final_pkgs
+    - store_artifacts:
+        path: C:/w/final_pkgs
+
+  binary_windows_test:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-medium-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    - checkout
+    - attach_workspace:
+        at: c:/users/circleci/project
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          script="/c/w/p/.circleci/scripts/binary_windows_test.sh"
+          cat "$script"
+          source "$script"
+
+  smoke_windows_test:
+    <<: *binary_windows_params
+    parameters:
+      build_environment:
+        type: string
+        default: ""
+      executor:
+        type: string
+        default: "windows-medium-cpu-with-nvidia-cuda"
+    executor: <<parameters.executor>>
+    steps:
+    - checkout
+    - run:
+        <<: *binary_checkout
+    - run:
+        <<: *binary_populate_env
+    - run:
+        name: Test
+        no_output_timeout: "1h"
+        command: |
+          set -eux -o pipefail
+          export TEST_NIGHTLY_PACKAGE=1
+          script="/c/w/p/.circleci/scripts/binary_windows_test.sh"
+          cat "$script"
+          source "$script"
+
   anaconda_prune:
     parameters:
       packages:

.circleci/verbatim-sources/job-specs/job-specs-custom.yml

@@ -24,6 +24,95 @@
           pushd /tmp/workspace
           git push -u origin "<< parameters.branch >>"
 
+  pytorch_python_doc_build:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-python-doc-push
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: large
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc Build and Push
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          # turn v1.12.0rc3 into 1.12
+          tag=$(echo $CIRCLE_TAG | sed -e 's/v*\([0-9]*\.[0-9]*\).*/\1/')
+          target=${tag:-main}
+          echo "building for ${target}"
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --env-file "${BASH_ENV}" --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && '"export CIRCLE_SHA1='$CIRCLE_SHA1'"' && . ./.circleci/scripts/python_doc_push_script.sh docs/'$target' '$target' site") | docker exec -u jenkins -i "$id" bash) 2>&1'
+
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts
+
+          mkdir -p ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/pytorch.github.io/docs/main ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/pytorch.github.io /tmp/workspace
+
+          # Save the docs build so we can debug any problems
+          export DEBUG_COMMIT_DOCKER_IMAGE=${COMMIT_DOCKER_IMAGE}-debug
+          docker commit "$id" ${DEBUG_COMMIT_DOCKER_IMAGE}
+          time docker push ${DEBUG_COMMIT_DOCKER_IMAGE}
+    - persist_to_workspace:
+        root: /tmp/workspace
+        paths:
+          - .
+    - store_artifacts:
+        path: ~/workspace/build_artifacts/main
+        destination: docs
+
+  pytorch_cpp_doc_build:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-cpp-doc-push
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: large
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc Build and Push
+        no_output_timeout: "1h"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          # turn v1.12.0rc3 into 1.12
+          tag=$(echo $CIRCLE_TAG | sed -e 's/v*\([0-9]*\.[0-9]*\).*/\1/')
+          target=${tag:-main}
+          echo "building for ${target}"
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --env-file "${BASH_ENV}" --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && '"export CIRCLE_SHA1='$CIRCLE_SHA1'"' && . ./.circleci/scripts/cpp_doc_push_script.sh docs/"$target" main") | docker exec -u jenkins -i "$id" bash) 2>&1'
+
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts
+
+          mkdir -p ~/workspace/build_artifacts
+          docker cp $id:/var/lib/jenkins/workspace/cppdocs/ /tmp/workspace
+
+          # Save the docs build so we can debug any problems
+          export DEBUG_COMMIT_DOCKER_IMAGE=${COMMIT_DOCKER_IMAGE}-debug
+          docker commit "$id" ${DEBUG_COMMIT_DOCKER_IMAGE}
+          time docker push ${DEBUG_COMMIT_DOCKER_IMAGE}
+
+    - persist_to_workspace:
+        root: /tmp/workspace
+        paths:
+          - .
+
   pytorch_macos_10_15_py3_build:
     environment:
       BUILD_ENVIRONMENT: pytorch-macos-10.15-py3-arm64-build
@@ -37,6 +126,7 @@
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export CROSS_COMPILE_ARM64=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
@@ -74,6 +164,7 @@
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             # Install sccache
@@ -95,198 +186,6 @@
           paths:
             - miniconda3
 
-  mac_build:
-    parameters:
-      build-environment:
-        type: string
-        description: Top-level label for what's being built/tested.
-      xcode-version:
-        type: string
-        default: "13.3.1"
-        description: What xcode version to build with.
-      build-generates-artifacts:
-        type: boolean
-        default: true
-        description: if the build generates build artifacts
-      python-version:
-        type: string
-        default: "3.8"
-    macos:
-      xcode: << parameters.xcode-version >>
-    resource_class: medium
-    environment:
-      BUILD_ENVIRONMENT: << parameters.build-environment >>
-      AWS_REGION: us-east-1
-    steps:
-
-      - checkout
-      - run_brew_for_macos_build
-
-      - run:
-          name: Install sccache
-          command: |
-            sudo curl --retry 3 https://s3.amazonaws.com/ossci-macos/sccache_v2.15 --output /usr/local/bin/sccache
-            sudo chmod +x /usr/local/bin/sccache
-            echo "export SCCACHE_BUCKET=ossci-compiler-cache-circleci-v2" >> "${BASH_ENV}"
-            echo "export SCCACHE_S3_KEY_PREFIX=${GITHUB_WORKFLOW}" >> "${BASH_ENV}"
-
-            set +x
-            echo "export AWS_ACCESS_KEY_ID=${CIRCLECI_AWS_ACCESS_KEY_FOR_SCCACHE_S3_BUCKET_V4}" >> "${BASH_ENV}"
-            echo "export AWS_SECRET_ACCESS_KEY=${CIRCLECI_AWS_SECRET_KEY_FOR_SCCACHE_S3_BUCKET_V4}" >> "${BASH_ENV}"
-            set -x
-
-      - run:
-          name: Get workflow job id
-          command: |
-            echo "export OUR_GITHUB_JOB_ID=${CIRCLE_WORKFLOW_JOB_ID}" >> "${BASH_ENV}"
-
-      - run:
-          name: Build
-          command: |
-            set -x
-
-            git submodule sync
-            git submodule update --init --recursive --depth 1 --jobs 0
-
-            export PATH="/usr/local/bin:$PATH"
-            export WORKSPACE_DIR="${HOME}/workspace"
-            mkdir -p "${WORKSPACE_DIR}"
-            MINICONDA_URL="https://repo.anaconda.com/miniconda/Miniconda3-py38_4.12.0-MacOSX-x86_64.sh"
-            if [  << parameters.python-version >> == 3.9.12 ]; then
-              MINICONDA_URL="https://repo.anaconda.com/miniconda/Miniconda3-py39_4.12.0-MacOSX-x86_64.sh"
-            fi
-
-            # If a local installation of conda doesn't exist, we download and install conda
-            if [ ! -d "${WORKSPACE_DIR}/miniconda3" ]; then
-              mkdir -p "${WORKSPACE_DIR}"
-              curl --retry 3 ${MINICONDA_URL} -o "${WORKSPACE_DIR}"/miniconda3.sh
-              bash "${WORKSPACE_DIR}"/miniconda3.sh -b -p "${WORKSPACE_DIR}"/miniconda3
-            fi
-            export PATH="${WORKSPACE_DIR}/miniconda3/bin:$PATH"
-            # shellcheck disable=SC1091
-            source "${WORKSPACE_DIR}"/miniconda3/bin/activate
-
-            brew link --force libomp
-
-            echo "export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname "$(which conda)")/../"}" >> "${BASH_ENV}"
-            .jenkins/pytorch/macos-build.sh
-
-      - when:
-          condition: << parameters.build-generates-artifacts >>
-          steps:
-            - run:
-                name: Archive artifacts into zip
-                command: |
-                  zip -1 -r artifacts.zip dist/ build/.ninja_log build/compile_commands.json .pytorch-test-times.json
-                  cp artifacts.zip /Users/distiller/workspace
-
-      - persist_to_workspace:
-          root: /Users/distiller/workspace/
-          paths:
-            - miniconda3
-            - artifacts.zip
-
-      - store_artifacts:
-          path: /Users/distiller/project/artifacts.zip
-
-  mac_test:
-    parameters:
-      build-environment:
-        type: string
-      shard-number:
-        type: string
-      num-test-shards:
-        type: string
-      xcode-version:
-        type: string
-      test-config:
-        type: string
-        default: 'default'
-
-    macos:
-      xcode: << parameters.xcode-version >>
-    environment:
-      GIT_DEFAULT_BRANCH: 'master'
-      BUILD_ENVIRONMENT: << parameters.build-environment >>
-      TEST_CONFIG: << parameters.test-config >>
-      SHARD_NUMBER: << parameters.shard-number >>
-      NUM_TEST_SHARDS: << parameters.num-test-shards >>
-      PYTORCH_RETRY_TEST_CASES: 1
-      PYTORCH_OVERRIDE_FLAKY_SIGNAL: 1
-    steps:
-      - checkout
-      - attach_workspace:
-          at: ~/workspace
-      - run_brew_for_macos_build
-      - run:
-          name: Test
-          no_output_timeout: "2h"
-          command: |
-            set -x
-
-            git submodule sync --recursive
-            git submodule update --init --recursive
-
-            mv ~/workspace/artifacts.zip .
-            unzip artifacts.zip
-
-            export IN_CI=1
-
-            COMMIT_MESSAGES=$(git cherry -v "origin/${GIT_DEFAULT_BRANCH:-master}")
-
-            export PATH="/usr/local/bin:$PATH"
-            export WORKSPACE_DIR="${HOME}/workspace"
-            mkdir -p "${WORKSPACE_DIR}"
-
-            export PATH="${WORKSPACE_DIR}/miniconda3/bin:$PATH"
-            source "${WORKSPACE_DIR}"/miniconda3/bin/activate
-
-            # sanitize the input commit message and PR body here:
-
-            # trim all new lines from commit messages to avoid issues with batch environment
-            # variable copying. see https://github.com/pytorch/pytorch/pull/80043#issuecomment-1167796028
-            COMMIT_MESSAGES="${COMMIT_MESSAGES//[$'\n\r']}"
-
-            # then trim all special characters like single and double quotes to avoid unescaped inputs to
-            # wreak havoc internally
-            export COMMIT_MESSAGES="${COMMIT_MESSAGES//[\'\"]}"
-
-            python3 -mpip install dist/*.whl
-            .jenkins/pytorch/macos-test.sh
-      - run:
-          name: Copy files for uploading test stats
-          command: |
-            # copy into a parent folder test-reports because we can't use CIRCLEI_BUILD_NUM in path when persisting to workspace
-            mkdir -p test-reports/test-reports_${CIRCLE_BUILD_NUM}/test/test-reports
-            cp -r test/test-reports test-reports/test-reports_${CIRCLE_BUILD_NUM}/test/test-reports
-      - store_test_results:
-          path: test/test-reports
-      - persist_to_workspace:
-          root: /Users/distiller/project/
-          paths:
-            - test-reports
-
-  upload_test_stats:
-    machine: # executor type
-      image: ubuntu-2004:202010-01 # # recommended linux image - includes Ubuntu 20.04, docker 19.03.13, docker-compose 1.27.4
-    steps:
-      - checkout
-      - attach_workspace:
-          at: ~/workspace
-      - run:
-          name: upload
-          command: |
-            set -ex
-            if [ -z ${AWS_ACCESS_KEY_FOR_OSSCI_ARTIFACT_UPLOAD} ]; then
-              echo "No credentials found, cannot upload test stats (are you on a fork?)"
-              exit 0
-            fi
-            cp -r ~/workspace/test-reports/* ~/project
-            pip3 install requests==2.26 rockset==0.8.3 boto3==1.19.12 six==1.16.0
-            export AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_FOR_OSSCI_ARTIFACT_UPLOAD}
-            export AWS_SECRET_ACCESS_KEY=${AWS_SECRET_KEY_FOR_OSSCI_ARTIFACT_UPLOAD}
-            # i dont know how to get the run attempt number for reruns so default to 1
-            python3 -m tools.stats.upload_test_stats --workflow-run-id "${CIRCLE_WORKFLOW_JOB_ID}" --workflow-run-attempt 1 --head-branch << pipeline.git.branch >> --circleci
   pytorch_macos_10_13_py3_test:
     environment:
       BUILD_ENVIRONMENT: pytorch-macos-10.13-py3-test
@@ -302,6 +201,7 @@
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             chmod a+x .jenkins/pytorch/macos-test.sh
@@ -314,6 +214,7 @@
             source /Users/distiller/workspace/miniconda3/bin/activate
             python3 -m pip install boto3==1.19.12
 
+            export IN_CI=1
             export JOB_BASE_NAME=$CIRCLE_JOB
 
             # Using the same IAM user to write stats to our OSS bucket
@@ -339,6 +240,7 @@
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             export BUILD_LITE_INTERPRETER=1
             export JOB_BASE_NAME=$CIRCLE_JOB
             chmod a+x ${HOME}/project/.jenkins/pytorch/macos-lite-interpreter-build-test.sh
@@ -428,6 +330,9 @@
           output_image=$docker_image_libtorch_android_x86_32-gradle
           docker commit "$id_x86_32" ${output_image}
           time docker push ${output_image}
+    - upload_binary_size_for_android_build:
+        build_type: prebuilt
+        artifacts: /home/circleci/workspace/build_android_artifacts/artifacts.tgz
     - store_artifacts:
         path: ~/workspace/build_android_artifacts/artifacts.tgz
         destination: artifacts.tgz
@@ -503,6 +408,9 @@
           output_image=${docker_image_libtorch_android_x86_32}-gradle
           docker commit "$id" ${output_image}
           time docker push ${output_image}
+    - upload_binary_size_for_android_build:
+        build_type: prebuilt-single
+        artifacts: /home/circleci/workspace/build_android_x86_32_artifacts/artifacts.tgz
     - store_artifacts:
         path: ~/workspace/build_android_x86_32_artifacts/artifacts.tgz
         destination: artifacts.tgz
@@ -512,43 +420,10 @@
     macos:
       xcode: "12.5.1"
     steps:
-      - run:
-          name: checkout with retry
-          command: |
-            checkout() {
-              set -ex
-              # Workaround old docker images with incorrect $HOME
-              # check https://github.com/docker/docker/issues/2968 for details
-              if [ "${HOME}" = "/" ]
-                then
-                export HOME=$(getent passwd $(id -un) | cut -d: -f6)
-              fi
-
-              mkdir -p ~/.ssh
-
-              echo 'github.com ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEAq2A7hRGmdnm9tUDbO9IDSwBK6TbQa+PXYPCPy6rbTrTtw7PHkccKrpp0yVhp5HdEIcKr6pLlVDBfOLX9QUsyCOV0wzfjIJNlGEYsdlLJizHhbn2mUjvSAHQqZETYP81eFzLQNnPHt4EVVUh7VfDESU84KezmD5QlWpXLmvU31/yMf+Se8xhHTvKSCZIFImWwoG6mbUoWf9nzpIoaSjB+weqqUUmpaaasXVal72J+UX2B+2RPW3RcT0eOzQgqlJL3RKrTJvdsjE3JEAvGq3lGHSZXy28G3skua2SmVi/w4yCE6gbODqnTWlg7+wC604ydGXA8VJiS5ap43JXiUFFAaQ==
-              ' >> ~/.ssh/known_hosts
-
-              # use git+ssh instead of https
-              git config --global url."ssh://git@github.com".insteadOf "https://github.com" || true
-              git config --global gc.auto 0 || true
-
-              echo 'Cloning git repository'
-              mkdir -p '/Users/distiller/project'
-              cd '/Users/distiller/project'
-              git clone "$CIRCLE_REPOSITORY_URL" .
-              echo 'Checking out branch'
-              git checkout --force -B "$CIRCLE_BRANCH" "$CIRCLE_SHA1"
-              git --no-pager log --no-color -n 1 --format='HEAD is now at %h %s'
-            }
-
-            retry () {
-              $* || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
-            }
-            retry checkout
+      - checkout
       - run_brew_for_ios_build
       - run:
-          name: Setup Fastlane
+          name: Run Fastlane
           no_output_timeout: "1h"
           command: |
             set -e
@@ -556,11 +431,26 @@
             cd ${PROJ_ROOT}/ios/TestApp
             # install fastlane
             sudo gem install bundler && bundle install
+            # install certificates
+            echo ${IOS_CERT_KEY_2022} >> cert.txt
+            base64 --decode cert.txt -o Certificates.p12
+            rm cert.txt
+            bundle exec fastlane install_root_cert
+            bundle exec fastlane install_dev_cert
+            # install the provisioning profile
+            PROFILE=PyTorch_CI_2022.mobileprovision
+            PROVISIONING_PROFILES=~/Library/MobileDevice/Provisioning\ Profiles
+            mkdir -pv "${PROVISIONING_PROFILES}"
+            cd "${PROVISIONING_PROFILES}"
+            echo ${IOS_SIGN_KEY_2022} >> cert.txt
+            base64 --decode cert.txt -o ${PROFILE}
+            rm cert.txt
       - run:
           name: Build
           no_output_timeout: "1h"
           command: |
             set -e
+            export IN_CI=1
             WORKSPACE=/Users/distiller/workspace
             PROJ_ROOT=/Users/distiller/project
             export TCLLIBPATH="/usr/local/lib"
@@ -613,12 +503,18 @@
           command: |
             set -e
             PROJ_ROOT=/Users/distiller/project
+            PROFILE=PyTorch_CI_2022
             # run the ruby build script
             if ! [ -x "$(command -v xcodebuild)" ]; then
               echo 'Error: xcodebuild is not installed.'
               exit 1
             fi
-            ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM}
+            echo ${IOS_DEV_TEAM_ID}
+            if [ ${IOS_PLATFORM} != "SIMULATOR" ]; then
+              ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM} -c ${PROFILE} -t ${IOS_DEV_TEAM_ID}
+            else
+              ruby ${PROJ_ROOT}/scripts/xcode_build.rb -i ${PROJ_ROOT}/build_ios/install -x ${PROJ_ROOT}/ios/TestApp/TestApp.xcodeproj -p ${IOS_PLATFORM}
+            fi
             if ! [ "$?" -eq "0" ]; then
               echo 'xcodebuild failed!'
               exit 1
@@ -641,13 +537,12 @@
             cd ${PROJ_ROOT}/ios/TestApp/benchmark
             mkdir -p ../models
             if [ ${USE_COREML_DELEGATE} == 1 ]; then
-              pip install coremltools==5.0b5 protobuf==3.20.1 six==1.16.0
+              pip install coremltools==5.0b5
+              pip install six
               python coreml_backend.py
             else
-              cd "${PROJ_ROOT}"
-              python test/mobile/model_test/gen_test_model.py ios-test
+              python trace_model.py
             fi
-            cd "${PROJ_ROOT}/ios/TestApp/benchmark"
             if [ ${BUILD_LITE_INTERPRETER} == 1 ]; then
               echo "Setting up the TestApp for LiteInterpreter"
               ruby setup.rb --lite 1
@@ -655,10 +550,10 @@
               echo "Setting up the TestApp for Full JIT"
               ruby setup.rb
             fi
-            cd "${PROJ_ROOT}/ios/TestApp"
-            # instruments -s -devices
-            if [ "${BUILD_LITE_INTERPRETER}" == 1 ]; then
-              if [ "${USE_COREML_DELEGATE}" == 1 ]; then
+            cd ${PROJ_ROOT}/ios/TestApp
+            instruments -s -devices
+            if [ ${BUILD_LITE_INTERPRETER} == 1 ]; then
+              if [ ${USE_COREML_DELEGATE} == 1 ]; then
                 fastlane scan --only_testing TestAppTests/TestAppTests/testCoreML
               else
                 fastlane scan --only_testing TestAppTests/TestAppTests/testLiteInterpreter
@@ -760,3 +655,27 @@
             set -e
             python3 -m pip install requests
             python3 ./.circleci/scripts/trigger_azure_pipeline.py
+
+  pytorch_doc_test:
+    environment:
+      BUILD_ENVIRONMENT: pytorch-doc-test
+      DOCKER_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/pytorch-linux-xenial-py3.7-gcc5.4"
+    resource_class: medium
+    machine:
+      image: ubuntu-2004:202104-01
+    steps:
+    - checkout
+    - calculate_docker_image_tag
+    - setup_linux_system_environment
+    - setup_ci_environment
+    - run:
+        name: Doc test
+        no_output_timeout: "30m"
+        command: |
+          set -ex
+          export COMMIT_DOCKER_IMAGE=${DOCKER_IMAGE}:build-${DOCKER_TAG}-${CIRCLE_SHA1}
+          echo "DOCKER_IMAGE: "${COMMIT_DOCKER_IMAGE}
+          time docker pull ${COMMIT_DOCKER_IMAGE} >/dev/null
+          export id=$(docker run --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -t -d -w /var/lib/jenkins ${COMMIT_DOCKER_IMAGE})
+          export COMMAND='((echo "sudo chown -R jenkins workspace && cd workspace && . ./.jenkins/pytorch/docs-test.sh") | docker exec -u jenkins -i "$id" bash) 2>&1'
+          echo ${COMMAND} > ./command.sh && unbuffer bash ./command.sh | ts

.flake8

@@ -22,9 +22,6 @@ exclude =
     ./docs/caffe2,
     ./docs/cpp/src,
     ./docs/src,
-    ./functorch/docs,
-    ./functorch/examples,
-    ./functorch/notebooks,
     ./scripts,
     ./test/generated_type_hints_smoketest.py,
     ./third_party,

.git-blame-ignore-revs

@@ -18,13 +18,7 @@ cc11aaaa60aadf28e3ec278bce26a42c1cd68a4f
 e3900d2ba5c9f91a24a9ce34520794c8366d5c54
 # 2021-04-21 Removed all unqualified `type: ignore`
 75024e228ca441290b6a1c2e564300ad507d7af6
-# 2021-04-30 [PyTorch] Autoformat c10
-44cc873fba5e5ffc4d4d4eef3bd370b653ce1ce1
 # 2021-05-14 Removed all versionless Python shebangs
 2e26976ad3b06ce95dd6afccfdbe124802edf28f
 # 2021-06-07 Strictly typed everything in `.github` and `tools`
 737d920b21db9b4292d056ee1329945990656304
-# 2022-06-09 Apply clang-format to ATen headers
-95b15c266baaf989ef7b6bbd7c23a2d90bacf687
-# 2022-06-11 [lint] autoformat test/cpp and torch/csrc
-30fb2c4abaaaa966999eab11674f25b18460e609

.github/ISSUE_TEMPLATE/ci-sev.md

@@ -5,8 +5,6 @@ about: Tracking incidents for PyTorch's CI infra.
 
 > NOTE: Remember to label this issue with "`ci: sev`"
 
-**MERGE BLOCKING** <!-- remove this line if you don't want this SEV to block merges -->
-
 ## Current Status
 *Status could be: preemptive, ongoing, mitigated, closed. Also tell people if they need to take action to fix it (i.e. rebase)*.
 

.github/actionlint.yaml

@@ -12,7 +12,5 @@ self-hosted-runner:
     - windows.8xlarge.nvidia.gpu
     - bm-runner
     - linux.rocm.gpu
-    - macos-m1-12
-    - macos-12-xl
     - macos-12
     - macos12.3-m1

.github/actions/build-android/action.yml

@@ -37,10 +37,12 @@ runs:
       shell: bash
       env:
         BRANCH: ${{ inputs.branch }}
+        JOB_BASE_NAME: ${{ inputs.build-environment }}-build-and-test
         BUILD_ENVIRONMENT: pytorch-linux-xenial-py3-clang5-android-ndk-r19c-${{ inputs.arch-for-build-env }}-build"
         AWS_DEFAULT_REGION: us-east-1
         PR_NUMBER: ${{ github.event.pull_request.number }}
         SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
+        CUSTOM_TEST_ARTIFACT_BUILD_DIR: build/custom_test_artifacts
         SCCACHE_BUCKET: ossci-compiler-cache-circleci-v2
         DOCKER_IMAGE: ${{ inputs.docker-image  }}
         MATRIX_ARCH: ${{ inputs.arch }}
@@ -50,12 +52,16 @@ runs:
         export container_name
         container_name=$(docker run \
           -e BUILD_ENVIRONMENT \
+          -e JOB_BASE_NAME \
           -e MAX_JOBS="$(nproc --ignore=2)" \
           -e AWS_DEFAULT_REGION \
+          -e IS_GHA \
           -e PR_NUMBER \
           -e SHA1 \
           -e BRANCH \
+          -e GITHUB_RUN_ID \
           -e SCCACHE_BUCKET \
+          -e CUSTOM_TEST_ARTIFACT_BUILD_DIR \
           -e SKIP_SCCACHE_INITIALIZATION=1 \
           --env-file="/tmp/github_env_${GITHUB_RUN_ID}" \
           --security-opt seccomp=unconfined \

.github/actions/calculate-docker-image/action.yml

@@ -17,16 +17,6 @@ inputs:
   pull:
     description: If set to any value, run `docker pull`` on the calculated image.
     required: false
-  skip_push:
-    description: If set to true value, skip will be pushed, default is to skip so that pushing will be explicit
-    required: false
-    default: "true"
-  force_push:
-    description: If set to any value, always run the push
-    required: false
-  push-ghcr-image:
-    description: If set to any value, push docker image to the ghcr.io.
-    required: false
 
 outputs:
   docker-image:
@@ -41,7 +31,7 @@ runs:
       id: calculate-tag
       env:
         IS_XLA: ${{ inputs.xla == 'true' && 'true' || '' }}
-        XLA_IMAGE_TAG: v0.4
+        XLA_IMAGE_TAG: v0.2
         DOCKER_IMAGE_BASE: 308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/${{ inputs.docker-image-name }}
       run: |
         if [ -n "${IS_XLA}" ]; then
@@ -63,7 +53,6 @@ runs:
         BASE_REVISION: ${{ github.event.pull_request.base.sha || github.sha }}
         DOCKER_IMAGE: ${{ steps.calculate-tag.outputs.docker-image }}
         DOCKER_TAG: ${{ steps.calculate-tag.outputs.docker-tag }}
-        DOCKER_FORCE_PUSH: ${{ inputs.force_push }}
       run: |
         set -x
         # Check if image already exists, if it does then skip building it
@@ -86,15 +75,9 @@ runs:
         PREVIOUS_DOCKER_TAG=$(git rev-parse "$MERGE_BASE:.circleci/docker")
         # If no image exists but the hash is the same as the previous hash then we should error out here
         if [[ "${PREVIOUS_DOCKER_TAG}" = "${DOCKER_TAG}" ]]; then
-          echo "WARNING: Something has gone wrong and the previous image isn't available for the merge-base of your branch"
-          echo "         Will re-build docker image to store in local cache, TTS may be longer"
-          # NOTE: DOCKER_FORCE_PUSH will always be set to true for docker-builds.yml
-          if [[ "${DOCKER_FORCE_PUSH}" != "true" ]]; then
-            # In order to avoid a stampeding herd of jobs trying to push all at once we set it to
-            # skip the push. If this is negatively affecting TTS across the board the suggestion
-            # should be to run the docker-builds.yml workflow to generate the correct docker builds
-            echo ::set-output name=skip_push::true
-          fi
+          echo "ERROR: Something has gone wrong and the previous image isn't available for the merge-base of your branch"
+          echo "       contact the PyTorch team to restore the original images"
+          exit 1
         fi
         echo ::set-output name=rebuild::yes
 
@@ -103,11 +86,7 @@ runs:
       env:
         IMAGE_NAME: ${{inputs.docker-image-name}}
         DOCKER_SKIP_S3_UPLOAD: "1"
-        # Skip push if we don't need it, or if specified in the inputs
-        DOCKER_SKIP_PUSH: ${{ steps.check.outputs.skip_push || inputs.skip_push }}
         DOCKER_TAG: ${{ steps.calculate-tag.outputs.docker-tag }}
-        PUSH_GHCR_IMAGE: ${{ inputs.push-ghcr-image }}
-        GHCR_PAT: ${{ env.GHCR_PAT }}
       working-directory: .circleci/docker
       shell: bash
       run: |

.github/actions/checkout-pytorch/action.yml

@@ -23,14 +23,11 @@ runs:
       env:
         NO_SUDO: ${{ inputs.no-sudo }}
       run: |
-        retry () {
-          $* || (sleep 1 && $*) || (sleep 2 && $*) || (sleep 4 && $*) || (sleep 8 && $*)
-        }
         echo "${GITHUB_WORKSPACE}"
         if [ -z "${NO_SUDO}" ]; then
-          retry sudo rm -rf "${GITHUB_WORKSPACE}"
+          sudo rm -rf "${GITHUB_WORKSPACE}"
         else
-          retry rm -rf "${GITHUB_WORKSPACE}"
+          rm -rf "${GITHUB_WORKSPACE}"
         fi
         mkdir "${GITHUB_WORKSPACE}"
 

.github/actions/download-build-artifacts/action.yml

@@ -15,7 +15,7 @@ runs:
   steps:
     - name: Download PyTorch Build Artifacts from S3
       if: ${{ !inputs.use-gha }}
-      uses: seemethere/download-artifact-s3@v4
+      uses: seemethere/download-artifact-s3@v3
       with:
         name: ${{ inputs.name }}
 

.github/actions/filter-test-configs/action.yml

@@ -1,60 +0,0 @@
-name: Filter test configs matrix
-
-description: |
-  Apply filter to the test configs matrix to keep only entries specified
-  by the PR test-config labels. If no test-config label is set, the same
-  test configs matrix is returned untouched.
-
-inputs:
-  github-token:
-    description: GITHUB_TOKEN
-    required: true
-  test-matrix:
-    required: true
-    type: string
-    description: JSON description of what test configs to run.
-
-outputs:
-  test-matrix:
-    description: The filtered test configs matrix.
-    value: ${{ steps.filter.outputs.test-matrix }}
-  is-test-matrix-empty:
-    description: True if the filtered test configs matrix is empty. False otherwise.
-    value: ${{ steps.filter.outputs.is-test-matrix-empty }}
-
-runs:
-  using: composite
-  steps:
-    - uses: nick-fields/retry@71062288b76e2b6214ebde0e673ce0de1755740a
-      name: Setup dependencies
-      env:
-        GITHUB_TOKEN: ${{ inputs.github-token }}
-      with:
-        shell: bash
-        timeout_minutes: 10
-        max_attempts: 5
-        retry_wait_seconds: 30
-        command: |
-          set -eux
-          python3 -m pip install requests==2.26.0 pyyaml==6.0
-
-    - name: Parse ref
-      shell: bash
-      id: parse-ref
-      run: .github/scripts/parse_ref.py
-
-    - name: Select all requested test configurations
-      shell: bash
-      env:
-        GITHUB_TOKEN: ${{ inputs.github-token }}
-      id: filter
-      run: |
-        .github/scripts/filter_test_configs.py \
-          --test-matrix "${{ inputs.test-matrix }}" \
-          --pr-number "${{ github.event.pull_request.number }}" \
-          --tag "${{ steps.parse-ref.outputs.tag }}"
-
-    - name: Print the filtered test matrix
-      shell: bash
-      run: |
-        echo "${{ steps.filter.outputs.test-matrix }}"

.github/actions/get-workflow-job-id/action.yml

@@ -15,7 +15,7 @@ outputs:
 runs:
   using: composite
   steps:
-    - uses: nick-fields/retry@7d4a37704547a311dbb66ebdf5b23ec19374a767
+    - uses: nick-fields/retry@71062288b76e2b6214ebde0e673ce0de1755740a
       id: get-job-id
       env:
         GITHUB_TOKEN: ${{ inputs.github-token }}
@@ -25,7 +25,7 @@ runs:
         max_attempts: 5
         retry_wait_seconds: 30
         command: |
-          set -eux
+          set -x
           python3 -m pip install requests==2.26.0
           GHA_WORKFLOW_JOB_ID=$(python3 .github/scripts/get_workflow_job_id.py "${GITHUB_RUN_ID}" "${RUNNER_NAME}")
           echo "::set-output name=job-id::${GHA_WORKFLOW_JOB_ID}"

.github/actions/pull-docker-image/action.yml

@@ -0,0 +1,19 @@
+name: Pull docker image
+
+description: pull a specific docker image
+
+inputs:
+  docker-image:
+    description: the image to pull
+    required: true
+
+runs:
+  using: composite
+  steps:
+    - name: Pull Docker image
+      shell: bash
+      env:
+        DOCKER_IMAGE: ${{ inputs.docker-image }}
+      run: |
+        retry () { "$@"  || (sleep 1 && "$@") || (sleep 2 && "$@") }
+        retry docker pull "${DOCKER_IMAGE}"

.github/actions/setup-linux/action.yml

@@ -44,5 +44,4 @@ runs:
     - name: Preserve github env variables for use in docker
       shell: bash
       run: |
-        env | grep '^GITHUB' >> "/tmp/github_env_${GITHUB_RUN_ID}"
-        env | grep '^CI' >> "/tmp/github_env_${GITHUB_RUN_ID}"
+        env | grep '^GITHUB' > "/tmp/github_env_${GITHUB_RUN_ID}"

.github/actions/setup-rocm/action.yml

@@ -49,8 +49,7 @@ runs:
     - name: Preserve github env variables for use in docker
       shell: bash
       run: |
-        env | grep '^GITHUB' >> "/tmp/github_env_${GITHUB_RUN_ID}"
-        env | grep '^CI' >> "/tmp/github_env_${GITHUB_RUN_ID}"
+        env | grep '^GITHUB' > "/tmp/github_env_${GITHUB_RUN_ID}"
 
     - name: ROCm set GPU_FLAG
       shell: bash

.github/actions/setup-ssh/action.yml

@@ -0,0 +1,17 @@
+name: Setup SSH
+
+description: Adds ssh keys for current user to machine
+
+inputs:
+  github-secret:
+    description: GitHub token
+    required: true
+
+runs:
+  using: composite
+  steps:
+    - name: "Enable SSH (Click me for login details)"
+      uses: seemethere/add-github-ssh-key@v1
+      with:
+        GITHUB_TOKEN: ${{ inputs.github-secret }}
+        activate-with-label: false

.github/actions/setup-win/action.yml

@@ -58,8 +58,3 @@ runs:
       uses: actions/setup-python@v2
       with:
         python-version: "3.x"
-        cache: pip
-        cache-dependency-path: |
-          **/requirements.txt
-          **/.circleci/docker/requirements-ci.txt
-          **/.github/requirements-gha-cache.txt

.github/actions/teardown-linux/action.yml

@@ -0,0 +1,28 @@
+name: Teardown Linux
+
+description: Stuff that should always run at the end of a linux job
+
+inputs:
+  skip-wait-ssh:
+    description: If set, don't wait for ssh to drain before tearing down
+    required: false
+    default: ""
+
+runs:
+  using: composite
+  steps:
+    - name: Hold runner for 2 hours or until ssh sessions have drained
+      # TODO working-directory: !{{ pytorch_directory }}
+      # Always hold for active ssh sessions
+      shell: bash
+      if: inputs.skip-wait-ssh == ''
+      run: .github/scripts/wait_for_ssh_to_drain.sh
+
+    - name: Kill containers, clean up images
+      shell: bash
+      run: |
+        # ignore expansion of "docker ps -q" since it could be empty
+        # shellcheck disable=SC2046
+        docker stop $(docker ps -q) || true
+        # Prune all of the docker images
+        docker system prune -af

.github/actions/teardown-rocm/action.yml

@@ -0,0 +1,25 @@
+name: Teardown ROCm host
+
+description: Teardown ROCm host for CI
+
+runs:
+  using: composite
+  steps:
+    - name: Kill containers, clean up images
+      if: always()
+      shell: bash
+      run: |
+        # ignore expansion of "docker ps -q" since it could be empty
+        # shellcheck disable=SC2046
+        docker stop $(docker ps -q) || true
+        # Prune all of the docker containers
+        docker container prune -f
+        # Prune everything docker if there are more than 10 images (~200GB).
+        # This is easier than using a time filter, e.g., "until=24h".
+        image_count=$(docker images | wc -l)
+        if [[ ${image_count} -gt 10 ]]; then
+            echo "Purging all docker caches"
+            docker system prune -af
+        else
+            echo "Will not purge docker, only ${image_count} images found"
+        fi

.github/actions/test-pytorch-binary/action.yml

@@ -1,41 +0,0 @@
-name: Test pytorch binary
-
-description: Pulls the docker image and tests the pytorch binary using it. All env variable referenced in the "Test PyTorch binary" step must be set in the GITHUB_ENV file
-
-runs:
-  using: composite
-  steps:
-    - name: Test PyTorch binary
-      shell: bash
-      run: |
-        set -x
-        # shellcheck disable=SC2086,SC2090
-        container_name=$(docker run \
-          ${GPU_FLAG:-} \
-          -e BINARY_ENV_FILE \
-          -e BUILDER_ROOT \
-          -e BUILD_ENVIRONMENT \
-          -e BUILD_SPLIT_CUDA \
-          -e DESIRED_CUDA \
-          -e DESIRED_DEVTOOLSET \
-          -e DESIRED_PYTHON \
-          -e GITHUB_ACTIONS \
-          -e GPU_ARCH_TYPE \
-          -e GPU_ARCH_VERSION \
-          -e LIBTORCH_VARIANT \
-          -e PACKAGE_TYPE \
-          -e PYTORCH_FINAL_PACKAGE_DIR \
-          -e PYTORCH_ROOT \
-          -e SKIP_ALL_TESTS \
-          --tty \
-          --detach \
-          -v "${GITHUB_WORKSPACE}/pytorch:/pytorch" \
-          -v "${GITHUB_WORKSPACE}/builder:/builder" \
-          -v "${RUNNER_TEMP}/artifacts:/final_pkgs" \
-          -w / \
-          "${DOCKER_IMAGE}"
-        )
-        docker exec -t -w "${PYTORCH_ROOT}" "${container_name}" bash -c "bash .circleci/scripts/binary_populate_env.sh"
-        # Generate test script
-        docker exec -t -w "${PYTORCH_ROOT}" -e OUTPUT_SCRIPT="/run.sh" "${container_name}" bash -c "bash .circleci/scripts/binary_linux_test.sh"
-        docker exec -t "${container_name}" bash -c "source ${BINARY_ENV_FILE} && bash -x /run.sh"

.github/actions/upload-test-artifacts/action.yml

@@ -36,20 +36,6 @@ runs:
         rm -f test-reports-*.zip
         zip -r "test-reports-${FILE_SUFFIX}.zip" test -i '*.xml'
 
-    - name: Zip usage log for upload
-      if: runner.os != 'Windows' && !inputs.use-gha
-      shell: bash
-      env:
-        FILE_SUFFIX: ${{ inputs.file-suffix }}
-      run: |
-        # Remove any previous test reports if they exist
-        rm -f usage-log-*.zip
-        # this workflow is also run in bazel build test, but we dont generate usage reports for it
-        # so check to see if the file exists first
-        if [ -f 'usage_log.txt' ]; then
-            zip "usage-log-${FILE_SUFFIX}.zip" 'usage_log.txt'
-        fi
-
     # Windows zip
     - name: Zip JSONs for upload
       if: runner.os == 'Windows' && !inputs.use-gha
@@ -69,46 +55,23 @@ runs:
         # -ir => recursive include all files in pattern
         7z a "test-reports-$Env:FILE_SUFFIX.zip" -ir'!test\*.xml'
 
-    - name: Zip usage log for upload
-      if: runner.os == 'Windows' && !inputs.use-gha
-      shell: powershell
-      env:
-        FILE_SUFFIX: ${{ inputs.file-suffix }}
-      run: |
-        # -ir => recursive include all files in pattern
-        7z a "usage-log-$Env:FILE_SUFFIX.zip" 'usage_log.txt'
-
     # S3 upload
     - name: Store Test Downloaded JSONs on S3
-      uses: seemethere/upload-artifact-s3@v5
+      uses: seemethere/upload-artifact-s3@v4
       if: ${{ !inputs.use-gha }}
       with:
-        s3-prefix: |
-          ${{ github.repository }}/${{ github.run_id }}/${{ github.run_attempt }}/artifact
         retention-days: 14
         if-no-files-found: warn
         path: test-jsons-*.zip
 
     - name: Store Test Reports on S3
-      uses: seemethere/upload-artifact-s3@v5
+      uses: seemethere/upload-artifact-s3@v4
       if: ${{ !inputs.use-gha }}
       with:
-        s3-prefix: |
-          ${{ github.repository }}/${{ github.run_id }}/${{ github.run_attempt }}/artifact
         retention-days: 14
         if-no-files-found: error
         path: test-reports-*.zip
 
-    - name: Store Usage Logs on S3
-      uses: seemethere/upload-artifact-s3@v5
-      if: ${{ !inputs.use-gha }}
-      with:
-        s3-prefix: |
-          ${{ github.repository }}/${{ github.run_id }}/${{ github.run_attempt }}/artifact
-        retention-days: 14
-        if-no-files-found: ignore
-        path: usage-log-*.zip
-
     # GHA upload
     - name: Store Test Downloaded JSONs on Github
       uses: actions/upload-artifact@v2

.github/auto_request_review.yml

@@ -1,27 +0,0 @@
-# Documented at https://github.com/necojackarc/auto-request-review
-reviewers:
-  groups:
-    symbolic-shapes:
-      - ezyang
-      - Chillee
-      - wconstab
-      - anjali411
-      - albanD
-      - Krovatkin
-      - miladm
-
-  per_author:
-    symbolic-shapes:
-      - symbolic-shapes
-      - antoniojkim
-
-files:
-  # none yet, TODO: migrate CODEOWNERS here
-
-options:
-  ignore_draft: true
-  ignored_keywords:
-    - DO NOT REVIEW
-  # Just manually setup a self-referential per_author rule if you
-  # want group assignment
-  enable_group_assignment: false

.github/ci_commit_pins/torchdynamo.txt

@@ -1 +0,0 @@
-6ead5cae0d1234aa64db06fe230ef56e12ec76fe

.github/ci_commit_pins/vision.txt

@@ -1 +0,0 @@
-d7d90f56117ce0955332846a5f90b8d1346c4c09

.github/ci_commit_pins/xla.txt

@@ -1 +0,0 @@
-f2b36df6a1a80137eff7644e6d0f4eeb7ff429d6

.github/merge_rules.json

@@ -0,0 +1,114 @@
+[
+   {
+      "name": "ONNX exporter",
+      "patterns": [
+         ".jenkins/caffe2/*",
+         "scripts/onnx/**",
+         "docs/source/onnx.rst",
+         "test/onnx/**",
+         "test/jit/test_export_modes.py",
+         "aten/src/ATen/core/interned_strings.h",
+         "tools/onnx/**",
+         "torch/_C/__init__.pyi.in",
+         "torch/csrc/jit/passes/onnx.*",
+         "torch/csrc/jit/passes/onnx/**",
+         "torch/csrc/jit/serialization/export.*",
+         "torch/csrc/jit/serialization/onnx.*",
+         "torch/csrc/onnx/**",
+         "torch/onnx/**"
+      ],
+      "approved_by": ["BowenBao", "garymm"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "NVFuser",
+      "patterns": [
+         "test/test_jit_cuda_fuser.py",
+         "torch/csrc/jit/codegen/fuser/cuda/**",
+         "torch/csrc/jit/codegen/cuda/**",
+         "benchmarks/cpp/nvfuser/**"
+      ],
+      "approved_by": ["csarofeen", "ngimel", "jjsjann123"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "OSS CI",
+      "patterns": [".github/**", ".circleci/**", ".jenkins/**", "scripts/**", "tools/**"],
+      "approved_by": ["ezyang", "pytorch/pytorch-dev-infra"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "Documentation",
+      "patterns": ["docs/**", "torch/*docs.py"],
+      "approved_by": ["mruberry", "ngimel", "janeyx99"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "Mobile",
+      "patterns": ["ios/**", "android/**", "test/mobile/**"],
+      "approved_by": ["linbinyu", "kit1980", "IvanKobzarev", "dreiss"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "Linear Algebra",
+      "patterns": [
+         "aten/src/ATen/native/cuda/linalg/**",
+         "aten/src/ATen/LinalgBackend.h",
+         "aten/src/ATen/native/**/*LinearAlgebra*",
+         "docs/source/linalg.rst",
+         "torch/linalg/**",
+         "torch/_linalg_utils.py",
+         "torch/**/python_linalg_functions.*",
+         "torch/**/linalg.h",
+         "tools/autograd/templates/python_linalg_functions.cpp",
+         "test/test_linalg.py"
+      ],
+      "approved_by": ["nikitaved", "mruberry", "pearu", "Lezcano", "IvanYashchuk"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "FFT",
+      "patterns": [
+         "aten/src/ATen/native/cuda/*FFT*.h",
+         "aten/src/ATen/native/SpectralOps.cpp",
+         "aten/src/ATen/native/mkl/SpectralOps.cpp",
+         "aten/src/ATen/native/cuda/SpectralOps.*",
+         "docs/source/fft.rst",
+         "torch/fft/**",
+         "torch/csrc/api/include/torch/fft.h",
+         "torch/**/python_fft_functions.*",
+         "tools/autograd/templates/python_fft_functions.cpp",
+         "test/cpp/api/fft.cpp"
+      ],
+      "approved_by": ["mruberry", "peterbell10"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "Sparse",
+      "patterns": [
+         "benchmarks/sparse",
+         "c10/util/sparse_bitset.h",
+         "docs/source/sparse.rst",
+         "torch/**/sparse/**",
+         "torch/**/*sparse*",
+         "torch/optim/sparse*",
+         "torch/ao/nn/sparse/**",
+         "torch/utils/benchmark/**/*sparse*",
+         "aten/src/ATen/native/ao_sparse/**",
+         "aten/src/ATen/native/sparse/**",
+         "aten/src/ATen/**/*Sparse*",
+         "aten/src/ATen/*Sparse*",
+         "torch/_masked/**",
+         "test/*_masked*",
+         "test/**/*sparse*"
+      ],
+      "approved_by": ["nikitaved", "cpuhrsch", "pearu", "IvanYashchuk"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   },
+   {
+      "name": "superuser",
+      "patterns": ["*"],
+      "approved_by": ["pytorch/metamates"],
+      "mandatory_checks_name": ["Facebook CLA Check", "Lint"]
+   }
+]

.github/merge_rules.yaml

@@ -1,356 +0,0 @@
-- name: ONNX exporter
-  patterns:
-  - .jenkins/caffe2/*
-  - aten/src/ATen/core/interned_strings.h
-  - docs/source/onnx.rst
-  - docs/source/onnx*
-  - docs/source/scripts/onnx/**
-  - scripts/onnx/**
-  - test/jit/test_export_modes.py
-  - test/onnx/**
-  - tools/onnx/**
-  - torch/_C/__init__.pyi.in
-  - torch/csrc/jit/passes/onnx.*
-  - torch/csrc/jit/passes/onnx/**
-  - torch/csrc/jit/serialization/export.*
-  - torch/csrc/jit/serialization/onnx.*
-  - torch/csrc/onnx/**
-  - torch/onnx/**
-  - third_party/onnx
-  - caffe2/python/onnx/**
-  approved_by:
-  - BowenBao
-  - abock
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: NVFuser
-  patterns:
-  - test/test_jit_cuda_fuser.py
-  - torch/csrc/jit/codegen/fuser/cuda/**
-  - torch/csrc/jit/codegen/cuda/**
-  - benchmarks/cpp/nvfuser/**
-  approved_by:
-  - csarofeen
-  - ngimel
-  - jjsjann123
-  - ptrblck
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: OSS CI
-  patterns:
-  - .github/**
-  - .circleci/**
-  - .jenkins/**
-  - scripts/**
-  - tools/**
-  approved_by:
-  - alband
-  - dagitses
-  - pytorch/pytorch-dev-infra
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: OSS CI / pytorchbot
-  patterns:
-  - .github/ci_commit_pins/vision.txt
-  - .github/ci_commit_pins/torchdynamo.txt
-  approved_by:
-  - pytorchbot
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: OSS CI / pytorchbot / XLA
-  patterns:
-  - .github/ci_commit_pins/xla.txt
-  approved_by:
-  - pytorchbot
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull / linux-bionic-py3_7-clang8-xla / build
-  - pull / linux-bionic-py3_7-clang8-xla / test (xla, 1, 1, linux.2xlarge)
-
-- name: Documentation
-  patterns:
-  - docs/**
-  - torch/*docs.py
-  approved_by:
-  - svekars
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Mobile
-  patterns:
-  - ios/**
-  - android/**
-  - test/mobile/**
-  approved_by:
-  - linbinyu
-  - IvanKobzarev
-  - dreiss
-  - raziel
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Linear Algebra
-  patterns:
-  - aten/src/ATen/native/cuda/linalg/**
-  - aten/src/ATen/LinalgBackend.h
-  - aten/src/ATen/native/**LinearAlgebra*
-  - docs/source/linalg.rst
-  - torch/linalg/**
-  - torch/_linalg_utils.py
-  - torch/**python_linalg_functions.*
-  - torch/**linalg.h
-  - tools/autograd/templates/python_linalg_functions.cpp
-  - test/test_linalg.py
-  approved_by:
-  - mruberry
-  - lezcano
-  - IvanYashchuk
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: FFT
-  patterns:
-  - aten/src/ATen/native/cuda/*FFT*.h
-  - aten/src/ATen/native/SpectralOps.cpp
-  - aten/src/ATen/native/mkl/SpectralOps.cpp
-  - aten/src/ATen/native/cuda/SpectralOps.*
-  - docs/source/fft.rst
-  - torch/fft/**
-  - torch/csrc/api/include/torch/fft.h
-  - torch/**python_fft_functions.*
-  - tools/autograd/templates/python_fft_functions.cpp
-  - test/cpp/api/fft.cpp
-  approved_by:
-  - mruberry
-  - peterbell10
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Sparse
-  patterns:
-  - benchmarks/sparse
-  - c10/util/sparse_bitset.h
-  - docs/source/sparse.rst
-  - torch/**sparse/**
-  - torch/**sparse*
-  - torch/optim/sparse*
-  - torch/ao/nn/sparse/**
-  - torch/utils/benchmark/**sparse*
-  - aten/src/ATen/native/ao_sparse/**
-  - aten/src/ATen/native/sparse/**
-  - aten/src/ATen/**Sparse*
-  - aten/src/ATen/*Sparse*
-  - torch/_masked/**
-  - test/*_masked*
-  - test/**sparse*
-  approved_by:
-  - nikitaved
-  - cpuhrsch
-  - pearu
-  - IvanYashchuk
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: MPS
-  patterns:
-  - test/test_mps.py
-  - aten/src/ATen/native/native_functions.yaml
-  - aten/src/ATen/mps/**
-  - aten/src/ATen/native/mps/**
-  approved_by:
-  - kulinseth
-  - alband
-  - malfet
-  - razarmehr
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-- name: Distributions
-  patterns:
-  - torch/distributions/**
-  - test/distributions/**
-  approved_by:
-  - fritzo
-  - neerajprad
-  - alicanb
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Distributed
-  patterns:
-  - docs/source/pipeline.rst
-  - docs/source/distributed*
-  - docs/source/rpc.rst
-  - docs/source/rpc/**
-  - docs/source/_static/img/rpc*
-  - docs/source/_static/img/*distributed*
-  - docs/source/elastic/**
-  - benchmarks/distributed/**
-  - torch/distributed/**
-  - torch/nn/parallel/distributed*
-  - torch/_C/_distributed*
-  - torch/csrc/distributed/**
-  - torch/testing/_internal/distributed/**
-  - test/distributed/**
-  - test/cpp/dist_autograd/**
-  - test/cpp/rpc/**
-  approved_by:
-  - mrshenli
-  - pritamdamania87
-  - zhaojuanmao
-  - rohan-varma
-  - wanchaol
-  - fduwjj
-  - H-Huang
-  - d4l3k
-  - aazzolini
-  - kwen2501
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: IDEEP
-  patterns:
-  - third_party/ideep
-  - caffe2/ideep/**
-  - caffe2/python/ideep/**
-  approved_by:
-  - XiaobingSuper
-  - yanbing-j
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: oneDNN graph
-  patterns:
-  - torch/csrc/jit/codegen/onednn/**
-  - test/test_jit_llga_fuser.py
-  approved_by:
-  - sanchitintel
-  - chunyuan-w
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: CPU ATen backend
-  patterns:
-  - aten/src/ATen/cpu/**
-  - aten/src/ATen/native/cpu/**
-  - aten/src/ATen/native/quantized/cpu/**
-  - aten/src/ATen/native/Convolution*.cpp
-  - aten/src/ATen/native/mkldnn/**
-  approved_by:
-  - mingfeima
-  - XiaobingSuper
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: CPU frontend
-  patterns:
-  - torch/cpu/**
-  - torch/utils/mkldnn.py
-  - test/test_mkldnn.py
-  approved_by:
-  - leslie-fang-intel
-  - CaoE
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Autocast
-  patterns:
-  - torch/amp/**
-  - aten/src/ATen/autocast_mode.*
-  - torch/csrc/jit/passes/autocast.cpp
-  - test/test_autocast.py
-  approved_by:
-  - leslie-fang-intel
-  - CaoE
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Lazy Tensor
-  patterns:
-  - torch/csrc/lazy/**
-  - test/cpp/lazy/**
-  - test/lazy/**
-  - codegen/api/lazy.py
-  - codegen/dest/lazy_ir.py
-  - codegen/dest/lazy_ts_lowering.py
-  - codegen/gen_lazy_tensor.py
-  - aten/src/ATen/native/ts_native_functions.yaml
-  approved_by:
-  - alanwaketan
-  - JackCaoG
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: superuser
-  patterns:
-  - '*'
-  approved_by:
-  - pytorch/metamates
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Core Reviewers
-  patterns:
-  - '*'
-  approved_by:
-  - mruberry
-  - lezcano
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull
-
-- name: Core Maintainers
-  patterns:
-  - '*'
-  approved_by:
-  - soumith
-  - gchanan
-  - ezyang
-  - dzhulgakov
-  mandatory_checks_name:
-  - EasyCLA
-  - Lint
-  - pull

.github/requirements-gha-cache.txt

@@ -1,16 +0,0 @@
-# This file is to cache other dependencies not specified elsewhere in:
-#   requirement.txt
-#   requirements-flake8.txt
-#   docs/requirements.txt
-#   docs/cpp/requirements.txt
-#   functorch/docs/requirements.txt
-#   .circleci/docker/requirements-ci.txt
-cffi==1.15.0
-dataclasses==0.6
-jinja2==3.0.1
-lintrunner=0.9.2
-ninja==1.10.0.post1
-pynvml==11.4.1
-requests==2.26
-rich==10.9.0
-rockset==0.8.10

.github/scale-config.yml

@@ -0,0 +1,69 @@
+# scale-config.yml:
+#   Powers what instance types are available for GHA auto-scaled
+#   runners. Runners listed here will be available as self hosted
+#   runners, configuration is directly pulled from the main branch.
+#
+# NOTE (Apr, 5, 2021): Linux runners are currently all an amazonlinux2
+#
+# NOTE (Jan 5, 2021): Linux runners are all non-ephemeral to reduce the amount of CreateInstaces calls
+#                     to avoid RequestLimitExceeded issues
+#
+# TODO: Add some documentation on how the auto-scaling works
+#
+# NOTE: Default values,
+#
+# runner_types:
+#   runner_label:
+#     instance_type: m4.large
+#     os: linux
+#     max_available: 20
+#     disk_size: 50
+#     is_ephemeral: true
+
+runner_types:
+  # mainly used for ciflow-should-run, not made to run any serious tests
+  linux.large:
+    instance_type: c5.large
+    os: linux
+    disk_size: 10
+    is_ephemeral: false
+  linux.2xlarge:
+    instance_type: c5.2xlarge
+    os: linux
+    max_available: 1000
+    disk_size: 150
+    is_ephemeral: false
+  linux.4xlarge: # for binary-builds
+    instance_type: c5.4xlarge
+    os: linux
+    max_available: 500
+    disk_size: 150
+    is_ephemeral: false
+  linux.8xlarge.nvidia.gpu:
+    instance_type: g3.8xlarge
+    os: linux
+    max_available: 200
+    disk_size: 150
+    is_ephemeral: false
+  linux.4xlarge.nvidia.gpu:
+    instance_type: g3.4xlarge
+    os: linux
+    max_available: 250
+    disk_size: 150
+    is_ephemeral: false
+  linux.16xlarge.nvidia.gpu:
+    instance_type: g3.16xlarge
+    os: linux
+    max_available: 10
+    disk_size: 150
+    is_ephemeral: false
+  windows.4xlarge:
+    instance_type: c5d.4xlarge
+    os: windows
+    max_available: 200
+    disk_size: 256
+  windows.8xlarge.nvidia.gpu:
+    instance_type: p3.2xlarge
+    os: windows
+    max_available: 50
+    disk_size: 256

.github/scripts/comment_on_pr.py

@@ -1,34 +0,0 @@
-from typing import Any
-from trymerge import gh_post_pr_comment
-from gitutils import get_git_remote_name, get_git_repo_dir, GitRepo
-from trymerge_explainer import BOT_COMMANDS_WIKI
-import os
-
-
-def parse_args() -> Any:
-    from argparse import ArgumentParser
-
-    parser = ArgumentParser("Comment on a PR")
-    parser.add_argument("pr_num", type=int)
-    parser.add_argument("action", type=str)
-    return parser.parse_args()
-
-
-def main() -> None:
-    args = parse_args()
-    repo = GitRepo(get_git_repo_dir(), get_git_remote_name(), debug=True)
-    org, project = repo.gh_owner_and_name()
-    run_url = os.environ.get("GH_RUN_URL")
-
-    job_link = f"[job]({run_url})" if run_url is not None else "job"
-    msg = (
-        f"The {args.action} {job_link} was canceled. If you believe this is a mistake,"
-        + f"then you can re trigger it through [pytorch-bot]({BOT_COMMANDS_WIKI})."
-    )
-
-    gh_post_pr_comment(org, project, args.pr_num, msg)
-    print(org, project, args.pr_num, msg)
-
-
-if __name__ == "__main__":
-    main()

.github/scripts/fetch_latest_green_commit.py

@@ -1,123 +0,0 @@
-import sys
-from typing import Any, Dict, List, NamedTuple, Tuple
-from gitutils import _check_output
-
-import rockset  # type: ignore[import]
-import os
-import re
-
-def eprint(msg: str) -> None:
-    print(msg, file=sys.stderr)
-
-class WorkflowCheck(NamedTuple):
-    workflowName: str
-    name: str
-    jobName: str
-    conclusion: str
-
-def get_latest_commits() -> List[str]:
-    latest_viable_commit = _check_output(
-        [
-            "git",
-            "log",
-            "-n",
-            "1",
-            "--pretty=format:%H",
-            "origin/viable/strict",
-        ],
-        encoding="ascii",
-    )
-    commits = _check_output(
-        [
-            "git",
-            "rev-list",
-            f"{latest_viable_commit}^..HEAD",
-            "--remotes=*origin/master",
-        ],
-        encoding="ascii",
-    ).splitlines()
-
-    return commits
-
-def query_commits(commits: List[str], qlambda: Any) -> Any:
-    params = rockset.ParamDict()
-    params['shas'] = ",".join(commits)
-    results = qlambda.execute(parameters=params)
-
-    return results
-
-def print_commit_status(commit: str, results: Dict[str, Any]) -> None:
-    print(commit)
-    for check in results['results']:
-        if check['sha'] == commit:
-            print(f"\t{check['conclusion']:>10}: {check['name']}")
-
-def get_commit_results(commit: str, results: Dict[str, Any]) -> List[Dict[str, Any]]:
-    workflow_checks = []
-    for check in results['results']:
-        if check['sha'] == commit:
-            workflow_checks.append(WorkflowCheck(
-                workflowName=check['workflowName'],
-                name=check['name'],
-                jobName=check['jobName'],
-                conclusion=check['conclusion'],
-            )._asdict())
-    return workflow_checks
-
-def isGreen(commit: str, results: Dict[str, Any]) -> Tuple[bool, str]:
-    workflow_checks = get_commit_results(commit, results)
-
-    regex = {
-        "pull": False,
-        "trunk": False,
-        "lint": False,
-        "linux-binary": False,
-        "windows-binary": False,
-    }
-
-    for check in workflow_checks:
-        workflowName = check['workflowName']
-        conclusion = check['conclusion']
-        for required_check in regex:
-            if re.match(required_check, workflowName, flags=re.IGNORECASE):
-                if conclusion not in ["success", "skipped"]:
-                    return (False, workflowName + " checks were not successful")
-                else:
-                    regex[required_check] = True
-        if workflowName in ["periodic", "docker-release-builds"] and conclusion not in ["success", "skipped"]:
-            return (False, workflowName + " checks were not successful")
-
-    missing_workflows = [x for x in regex.keys() if not regex[x]]
-    if len(missing_workflows) > 0:
-        return (False, "missing required workflows: " + ", ".join(missing_workflows))
-
-    return (True, "")
-
-def get_latest_green_commit(commits: List[str], results: Dict[str, Any]) -> Any:
-    for commit in commits:
-        eprint(f"Checking {commit}")
-        is_green, msg = isGreen(commit, results)
-        if is_green:
-            eprint("GREEN")
-            return commit
-        else:
-            eprint("RED: " + msg)
-    return None
-
-def main() -> None:
-    rs = rockset.Client(
-        api_server="api.rs2.usw2.rockset.com", api_key=os.environ["ROCKSET_API_KEY"]
-    )
-    qlambda = rs.QueryLambda.retrieve(
-        'commit_jobs_batch_query',
-        version='15aba20837ae9d75',
-        workspace='commons')
-
-    commits = get_latest_commits()
-    results = query_commits(commits, qlambda)
-
-    latest_viable_commit = get_latest_green_commit(commits, results)
-    print(latest_viable_commit)
-
-if __name__ == "__main__":
-    main()

.github/scripts/filter_test_configs.py

@@ -1,162 +0,0 @@
-#!/usr/bin/env python3
-
-import sys
-import re
-import json
-import os
-import requests
-from typing import Any, Dict, Set, List
-import yaml
-import warnings
-
-PREFIX = "test-config/"
-
-# Same as shard names
-VALID_TEST_CONFIG_LABELS = {f"{PREFIX}{label}" for label in {
-    "backwards_compat",
-    "crossref",
-    "default",
-    "deploy",
-    "distributed",
-    "docs_tests",
-    "dynamo",
-    "force_on_cpu",
-    "functorch",
-    "jit_legacy",
-    "multigpu",
-    "nogpu_AVX512",
-    "nogpu_NO_AVX2",
-    "slow",
-    "xla",
-}}
-
-def parse_args() -> Any:
-    from argparse import ArgumentParser
-    parser = ArgumentParser("Filter all test configurations and keep only requested ones")
-    parser.add_argument("--test-matrix", type=str, required=True, help="the original test matrix")
-    parser.add_argument("--pr-number", type=str, help="the pull request number")
-    parser.add_argument("--tag", type=str, help="the associated tag if it exists")
-    return parser.parse_args()
-
-
-def get_labels(pr_number: int) -> Set[str]:
-    """
-    Dynamical get the latest list of labels from the pull request
-    """
-    # From https://docs.github.com/en/actions/learn-github-actions/environment-variables
-    PYTORCH_REPO = os.environ.get("GITHUB_REPOSITORY", "pytorch/pytorch")
-    PYTORCH_GITHUB_API = f"https://api.github.com/repos/{PYTORCH_REPO}"
-    GITHUB_TOKEN = os.environ["GITHUB_TOKEN"]
-
-    REQUEST_HEADERS = {
-        "Accept": "application/vnd.github.v3+json",
-        "Authorization": "token " + GITHUB_TOKEN,
-    }
-
-    response = requests.get(
-        f"{PYTORCH_GITHUB_API}/issues/{pr_number}/labels",
-        headers=REQUEST_HEADERS,
-    )
-
-    if response.status_code != requests.codes.ok:
-        warnings.warn(f"Failed to get the labels for #{pr_number} (status code {response.status_code})")
-        return set()
-
-    return {label.get("name") for label in response.json() if label.get("name")}
-
-
-def filter(test_matrix: Dict[str, List[Any]], labels: Set[str]) -> Dict[str, List[Any]]:
-    """
-    Select the list of test config to run from the test matrix. The logic works
-    as follows:
-
-    If the PR has one or more labels as specified in the VALID_TEST_CONFIG_LABELS set, only
-    these test configs will be selected.  This also works with ciflow labels, for example,
-    if a PR has both ciflow/trunk and test-config/functorch, only trunk functorch builds
-    and tests will be run
-
-    If the PR has none of the test-config label, all tests are run as usual.
-    """
-
-    filtered_test_matrix: Dict[str, List[Any]] = {
-        "include": []
-    }
-
-    for entry in test_matrix.get("include", []):
-        config_name = entry.get("config", "")
-        if not config_name:
-            continue
-
-        label = f"{PREFIX}{config_name.strip()}"
-        if label in labels:
-            print(f"Select {config_name} because label {label} is presented in the pull request by the time the test starts")
-            filtered_test_matrix["include"].append(entry)
-
-    valid_test_config_labels = labels.intersection(VALID_TEST_CONFIG_LABELS)
-
-    if not filtered_test_matrix["include"] and not valid_test_config_labels:
-        # Found no valid label and the filtered test matrix is empty, return the same
-        # test matrix as before so that all tests can be run normally
-        return test_matrix
-    else:
-        # When the filter test matrix contain matches or if a valid test config label
-        # is found in the PR, return the filtered test matrix
-        return filtered_test_matrix
-
-
-def main() -> None:
-    args = parse_args()
-    # Load the original test matrix set by the workflow. Its format, however,
-    # doesn't follow the strict JSON format, so we load it using yaml here for
-    # its more relaxed syntax
-    test_matrix = yaml.safe_load(args.test_matrix)
-
-    if test_matrix is None:
-        warnings.warn(f"Invalid test matrix input '{args.test_matrix}', exiting")
-        # We handle invalid test matrix gracefully by marking it as empty
-        print("::set-output name=is-test-matrix-empty::True")
-        sys.exit(0)
-
-    pr_number = args.pr_number
-    tag = args.tag
-
-    # If the tag matches, we can get the PR number from it, this is from ciflow
-    # workflow dispatcher
-    tag_regex = re.compile(r"^ciflow/\w+/(?P<pr_number>\d+)$")
-
-    if pr_number:
-        # If a PR number is set, query all the labels from that PR
-        labels = get_labels(int(pr_number))
-        # Then filter the test matrix and keep only the selected ones
-        filtered_test_matrix = filter(test_matrix, labels)
-
-    elif tag:
-        m = tag_regex.match(tag)
-
-        if m:
-            pr_number = m.group("pr_number")
-
-            # The PR number can also come from the tag in ciflow tag event
-            labels = get_labels(int(pr_number))
-            # Filter the test matrix and keep only the selected ones
-            filtered_test_matrix = filter(test_matrix, labels)
-
-        else:
-            # There is a tag but it isn't ciflow, so there is nothing left to do
-            filtered_test_matrix = test_matrix
-
-    else:
-        # No PR number, no tag, we can just return the test matrix as it is
-        filtered_test_matrix = test_matrix
-
-    # Set the filtered test matrix as the output
-    print(f"::set-output name=test-matrix::{json.dumps(filtered_test_matrix)}")
-
-    filtered_test_matrix_len = len(filtered_test_matrix.get("include", []))
-    # and also put a flag if the test matrix is empty, so subsequent jobs can
-    # quickly check it without the need to parse the JSON string
-    print(f"::set-output name=is-test-matrix-empty::{filtered_test_matrix_len == 0}")
-
-
-if __name__ == "__main__":
-    main()

.github/scripts/generate_binary_build_matrix.py

@@ -13,10 +13,10 @@ architectures:
 from typing import Dict, List, Tuple, Optional
 
 
-CUDA_ARCHES = ["11.6", "11.7"]
+CUDA_ARCHES = ["10.2", "11.3", "11.6"]
 
 
-ROCM_ARCHES = ["5.1.1", "5.2"]
+ROCM_ARCHES = ["5.0", "5.1.1"]
 
 
 def arch_type(arch_version: str) -> str:
@@ -90,8 +90,11 @@ def generate_conda_matrix(os: str) -> List[Dict[str, str]]:
     ret: List[Dict[str, str]] = []
     arches = ["cpu"]
     python_versions = FULL_PYTHON_VERSIONS
-    if os == "linux" or os == "windows":
+    if os == "linux":
         arches += CUDA_ARCHES
+    elif os == "windows":
+        # We don't build CUDA 10.2 for window see https://github.com/pytorch/pytorch/issues/65648
+        arches += list_without(CUDA_ARCHES, ["10.2"])
     elif os == "macos-arm64":
         python_versions = list_without(python_versions, ["3.7"])
     for python_version in python_versions:
@@ -126,7 +129,8 @@ def generate_libtorch_matrix(os: str, abi_version: str,
             arches += CUDA_ARCHES
             arches += ROCM_ARCHES
         elif os == "windows":
-            arches += CUDA_ARCHES
+            # We don't build CUDA 10.2 for window see https://github.com/pytorch/pytorch/issues/65648
+            arches += list_without(CUDA_ARCHES, ["10.2"])
 
     if libtorch_variants is None:
         libtorch_variants = [
@@ -147,45 +151,25 @@ def generate_libtorch_matrix(os: str, abi_version: str,
             # ROCm builds without-deps failed even in ROCm runners; skip for now
             if gpu_arch_type == "rocm" and "without-deps" in libtorch_variant:
                 continue
-            desired_cuda = translate_desired_cuda(gpu_arch_type, gpu_arch_version)
-
-            if desired_cuda == "rocm5.1.1" and abi_version == PRE_CXX11_ABI:
-                ret.append(
-                    {
-                        "gpu_arch_type": gpu_arch_type,
-                        "gpu_arch_version": gpu_arch_version,
-                        "desired_cuda": desired_cuda,
-                        "libtorch_variant": libtorch_variant,
-                        "libtorch_config": abi_version if os == "windows" else "",
-                        "devtoolset": abi_version if os != "windows" else "",
-                        "container_image": (
-                            "pytorch/manylinux-builder:rocm5.1.1-cd2573d54f9bd9b8f32b4dd7f182923a846597d5"
-                            if os != "windows" else ""
-                        ),
-                        "package_type": "libtorch",
-                        "build_name": f"libtorch-{gpu_arch_type}{gpu_arch_version}-{libtorch_variant}-{abi_version}".replace(
-                            ".", "_"
-                        ),
-                    }
-                )
-            else:
-                ret.append(
-                    {
-                        "gpu_arch_type": gpu_arch_type,
-                        "gpu_arch_version": gpu_arch_version,
-                        "desired_cuda": desired_cuda,
-                        "libtorch_variant": libtorch_variant,
-                        "libtorch_config": abi_version if os == "windows" else "",
-                        "devtoolset": abi_version if os != "windows" else "",
-                        "container_image": LIBTORCH_CONTAINER_IMAGES[
-                            (arch_version, abi_version)
-                        ] if os != "windows" else "",
-                        "package_type": "libtorch",
-                        "build_name": f"libtorch-{gpu_arch_type}{gpu_arch_version}-{libtorch_variant}-{abi_version}".replace(
-                            ".", "_"
-                        ),
-                    }
-                )
+            ret.append(
+                {
+                    "gpu_arch_type": gpu_arch_type,
+                    "gpu_arch_version": gpu_arch_version,
+                    "desired_cuda": translate_desired_cuda(
+                        gpu_arch_type, gpu_arch_version
+                    ),
+                    "libtorch_variant": libtorch_variant,
+                    "libtorch_config": abi_version if os == "windows" else "",
+                    "devtoolset": abi_version if os != "windows" else "",
+                    "container_image": LIBTORCH_CONTAINER_IMAGES[
+                        (arch_version, abi_version)
+                    ] if os != "windows" else "",
+                    "package_type": "libtorch",
+                    "build_name": f"libtorch-{gpu_arch_type}{gpu_arch_version}-{libtorch_variant}-{abi_version}".replace(
+                        ".", "_"
+                    ),
+                }
+            )
     return ret
 
 
@@ -199,81 +183,37 @@ def generate_wheels_matrix(os: str,
 
     if python_versions is None:
         # Define default python version
-        python_versions = list(FULL_PYTHON_VERSIONS)
+        python_versions = FULL_PYTHON_VERSIONS
         if os == "macos-arm64":
             python_versions = list_without(python_versions, ["3.7"])
 
-        if os == "linux":
-            # NOTE: We only build 3.11 wheel on linux as 3.11 is not
-            # available on conda right now
-            python_versions.append("3.11")
-
     if arches is None:
         # Define default compute archivectures
         arches = ["cpu"]
         if os == "linux":
             arches += CUDA_ARCHES + ROCM_ARCHES
         elif os == "windows":
-            arches += CUDA_ARCHES
+            # We don't build CUDA 10.2 for window see https://github.com/pytorch/pytorch/issues/65648
+            arches += list_without(CUDA_ARCHES, ["10.2"])
 
     ret: List[Dict[str, str]] = []
     for python_version in python_versions:
         for arch_version in arches:
             gpu_arch_type = arch_type(arch_version)
             gpu_arch_version = "" if arch_version == "cpu" else arch_version
-            # Skip rocm 3.11 binaries for now as the docker image are not correct
-            if python_version == "3.11" and gpu_arch_type == "rocm":
-                continue
-
-            desired_cuda = translate_desired_cuda(gpu_arch_type, gpu_arch_version)
-            # special 11.7 wheels package without dependencies
-            # dependency downloaded via pip install
-            if arch_version == "11.7" and os == "linux":
-                ret.append(
-                    {
-                        "python_version": python_version,
-                        "gpu_arch_type": gpu_arch_type,
-                        "gpu_arch_version": gpu_arch_version,
-                        "desired_cuda": desired_cuda,
-                        "container_image": WHEEL_CONTAINER_IMAGES[arch_version],
-                        "package_type": package_type,
-                        "pytorch_extra_install_requirements":
-                        "nvidia-cuda-runtime-cu11; platform_system == 'Linux' | "
-                        "nvidia-cudnn-cu11==8.5.0.96; platform_system == 'Linux' | "
-                        "nvidia-cublas-cu11==11.10.3.66; platform_system == 'Linux'",
-                        "build_name":
-                        f"{package_type}-py{python_version}-{gpu_arch_type}{gpu_arch_version}-with-pypi-cudnn"
-                        .replace(
-                            ".", "_"
-                        ),
-                    }
-                )
-            if desired_cuda == "rocm5.1.1":
-                ret.append(
-                    {
-                        "python_version": python_version,
-                        "gpu_arch_type": gpu_arch_type,
-                        "gpu_arch_version": gpu_arch_version,
-                        "desired_cuda": desired_cuda,
-                        "container_image": "pytorch/manylinux-builder:rocm5.1.1-cd2573d54f9bd9b8f32b4dd7f182923a846597d5",
-                        "package_type": package_type,
-                        "build_name": f"{package_type}-py{python_version}-{gpu_arch_type}{gpu_arch_version}".replace(
-                            ".", "_"
-                        ),
-                    }
-                )
-            else:
-                ret.append(
-                    {
-                        "python_version": python_version,
-                        "gpu_arch_type": gpu_arch_type,
-                        "gpu_arch_version": gpu_arch_version,
-                        "desired_cuda": desired_cuda,
-                        "container_image": WHEEL_CONTAINER_IMAGES[arch_version],
-                        "package_type": package_type,
-                        "build_name": f"{package_type}-py{python_version}-{gpu_arch_type}{gpu_arch_version}".replace(
-                            ".", "_"
-                        ),
-                    }
-                )
+            ret.append(
+                {
+                    "python_version": python_version,
+                    "gpu_arch_type": gpu_arch_type,
+                    "gpu_arch_version": gpu_arch_version,
+                    "desired_cuda": translate_desired_cuda(
+                        gpu_arch_type, gpu_arch_version
+                    ),
+                    "container_image": WHEEL_CONTAINER_IMAGES[arch_version],
+                    "package_type": package_type,
+                    "build_name": f"{package_type}-py{python_version}-{gpu_arch_type}{gpu_arch_version}".replace(
+                        ".", "_"
+                    ),
+                }
+            )
     return ret

.github/scripts/generate_ci_workflows.py

@@ -17,6 +17,7 @@ Arch = Literal["windows", "linux", "macos"]
 GITHUB_DIR = Path(__file__).resolve().parent.parent
 
 LABEL_CIFLOW_TRUNK = "ciflow/trunk"
+LABEL_CIFLOW_ALL = "ciflow/all"
 LABEL_CIFLOW_BINARIES = "ciflow/binaries"
 LABEL_CIFLOW_PERIODIC = "ciflow/periodic"
 LABEL_CIFLOW_BINARIES_LIBTORCH = "ciflow/binaries_libtorch"
@@ -33,6 +34,7 @@ class CIFlowConfig:
 
     def __post_init__(self) -> None:
         if not self.isolated_workflow:
+            self.labels.add(LABEL_CIFLOW_ALL)
             if LABEL_CIFLOW_PERIODIC not in self.labels:
                 self.labels.add(LABEL_CIFLOW_TRUNK)
 
@@ -134,7 +136,7 @@ LINUX_BINARY_SMOKE_WORKFLOWS = [
         package_type="manywheel",
         build_configs=generate_binary_build_matrix.generate_wheels_matrix(
             OperatingSystem.LINUX,
-            arches=["11.6"],
+            arches=["10.2"],
             python_versions=["3.7"]),
         branches="master",
     ),
@@ -207,6 +209,15 @@ WINDOWS_BINARY_BUILD_WORKFLOWS = [
     ),
 ]
 WINDOWS_BINARY_SMOKE_WORKFLOWS = [
+    BinaryBuildWorkflow(
+        os=OperatingSystem.WINDOWS,
+        package_type="wheel",
+        build_configs=generate_binary_build_matrix.generate_wheels_matrix(
+            OperatingSystem.WINDOWS,
+            arches=["11.3"],
+            python_versions=["3.7"]),
+        branches="master",
+    ),
     BinaryBuildWorkflow(
         os=OperatingSystem.WINDOWS,
         package_type="libtorch",

.github/scripts/get_workflow_job_id.py

@@ -31,9 +31,7 @@ parser.add_argument(
 args = parser.parse_args()
 
 
-# From https://docs.github.com/en/actions/learn-github-actions/environment-variables
-PYTORCH_REPO = os.environ.get("GITHUB_REPOSITORY", "pytorch/pytorch")
-PYTORCH_GITHUB_API = f"https://api.github.com/repos/{PYTORCH_REPO}"
+PYTORCH_REPO = "https://api.github.com/repos/pytorch/pytorch"
 GITHUB_TOKEN = os.environ["GITHUB_TOKEN"]
 REQUEST_HEADERS = {
     "Accept": "application/vnd.github.v3+json",
@@ -41,7 +39,7 @@ REQUEST_HEADERS = {
 }
 
 response = requests.get(
-    f"{PYTORCH_GITHUB_API}/actions/runs/{args.workflow_run_id}/jobs?per_page=100",
+    f"{PYTORCH_REPO}/actions/runs/{args.workflow_run_id}/jobs?per_page=100",
     headers=REQUEST_HEADERS,
 )
 

.github/scripts/gitutils.py

@@ -305,8 +305,8 @@ def patterns_to_regex(allowed_patterns: List[str]) -> Any:
     """
     pattern is glob-like, i.e. the only special sequences it has are:
       - ? - matches single character
-      - * - matches any non-folder separator characters or no character
-      - ** - matches any characters or no character
+      - * - matches any non-folder separator characters
+      - ** - matches any characters
       Assuming that patterns are free of braces and backslashes
       the only character that needs to be escaped are dot and plus
     """
@@ -324,9 +324,9 @@ def patterns_to_regex(allowed_patterns: List[str]) -> Any:
             elif c == "*":
                 if pattern_.peek() == "*":
                     next(pattern_)
-                    rc += ".*"
+                    rc += ".+"
                 else:
-                    rc += "[^/]*"
+                    rc += "[^/]+"
             else:
                 rc += c
     rc += ")"

.github/scripts/install_nvidia_utils_linux.sh

@@ -2,10 +2,8 @@
 
 set -eou pipefail
 
-
-DISTRIBUTION=$(. /etc/os-release;echo $ID$VERSION_ID)
-DRIVER_VERSION="515.57"
-DRIVER_FN="NVIDIA-Linux-x86_64-${DRIVER_VERSION}.run"
+DISTRIBUTION=$(. /etc/os-release;echo $ID$VERSION_ID) \
+DRIVER_FN="NVIDIA-Linux-x86_64-510.60.02.run"
 YUM_REPO_URL="https://nvidia.github.io/nvidia-docker/${DISTRIBUTION}/nvidia-docker.repo"
 
 install_nvidia_docker2_amzn2() {
@@ -22,52 +20,17 @@ install_nvidia_docker2_amzn2() {
 install_nvidia_driver_amzn2() {
     (
         set -x
-
-        HAS_NVIDIA_DRIVER=0
-        # Check if NVIDIA driver has already been installed
-        if [ -x "$(command -v nvidia-smi)" ]; then
-            # The driver exists, check its version next
-            INSTALLED_DRIVER_VERSION=$(nvidia-smi --query-gpu=driver_version --format=csv,noheader)
-
-            if [ "$INSTALLED_DRIVER_VERSION" != "$DRIVER_VERSION" ]; then
-                # TODO
-                # Remove this after torchrec and FBGEMM have both been updated to use
-                # PyTorch NVIDIA installation script instead of using the latest driver
-                # from RHEL repo
-                HAS_NVIDIA_DRIVER=1
-                echo "NVIDIA driver ($INSTALLED_DRIVER_VERSION) has been installed, but we expect to have $DRIVER_VERSION instead. Skipping NVIDIA driver installation for now until torchrec and FBGEMM are updated to use PyTorch NVIDIA installation script instead of RHEL repo"
-            else
-                HAS_NVIDIA_DRIVER=1
-                echo "NVIDIA driver ($INSTALLED_DRIVER_VERSION) has already been installed. Skipping NVIDIA driver installation"
-            fi
-        fi
-
-        if [ "$HAS_NVIDIA_DRIVER" -eq 0 ]; then
-            sudo yum groupinstall -y "Development Tools"
-            # ensure our kernel install is the same as our underlying kernel,
-            # groupinstall "Development Tools" has a habit of mismatching kernel headers
-            sudo yum install -y "kernel-devel-uname-r == $(uname -r)"
-            sudo modprobe backlight
-            sudo curl -fsL -o /tmp/nvidia_driver "https://s3.amazonaws.com/ossci-linux/nvidia_driver/$DRIVER_FN"
-            sudo /bin/bash /tmp/nvidia_driver -s --no-drm || (sudo cat /var/log/nvidia-installer.log && false)
-            sudo rm -fv /tmp/nvidia_driver
-        fi
-
+        sudo yum groupinstall -y "Development Tools"
+        # ensure our kernel install is the same as our underlying kernel,
+        # groupinstall "Development Tools" has a habit of mismatching kernel headers
+        sudo yum install -y "kernel-devel-uname-r == $(uname -r)"
+        sudo curl -fsL -o /tmp/nvidia_driver "https://s3.amazonaws.com/ossci-linux/nvidia_driver/$DRIVER_FN"
+        sudo /bin/bash /tmp/nvidia_driver -s --no-drm || (sudo cat /var/log/nvidia-installer.log && false)
+        sudo rm -fv /tmp/nvidia_driver
         nvidia-smi
     )
 }
 
-echo "== Installing nvidia driver ${DRIVER_FN} =="
-case "${DISTRIBUTION}" in
-    amzn*)
-        install_nvidia_driver_amzn2
-        ;;
-    *)
-        echo "ERROR: Unknown distribution ${DISTRIBUTION}"
-        exit 1
-        ;;
-esac
-
 # Install container toolkit based on distribution
 echo "== Installing nvidia container toolkit for ${DISTRIBUTION} =="
 case "${DISTRIBUTION}" in
@@ -79,3 +42,14 @@ case "${DISTRIBUTION}" in
         exit 1
         ;;
 esac
+
+echo "== Installing nvidia driver ${DRIVER_FN} =="
+case "${DISTRIBUTION}" in
+    amzn*)
+        install_nvidia_driver_amzn2
+        ;;
+    *)
+        echo "ERROR: Unknown distribution ${DISTRIBUTION}"
+        exit 1
+        ;;
+esac

.github/scripts/pr-sanity-check.sh

@@ -1,60 +0,0 @@
-#!/usr/bin/env bash
-
-set -eou pipefail
-
-GIT_TOP_DIR=$(git rev-parse --show-toplevel)
-
-TMPFILE=$(mktemp)
-trap "rm -rf ${TMPFILE}" EXIT
-
-# By default just run against the latest commit
-BASE=${BASE:-HEAD~1}
-HEAD=${HEAD:-HEAD}
-
-ancestor=$(git merge-base "${BASE}" "${HEAD}")
-echo "INFO: Checking aginst the following stats"
-(
-    set -x
-    git diff --stat "$ancestor" "${HEAD}" | sed '$d' > "${TMPFILE}"
-)
-
-while read -r git_attribute; do
-    if echo "${git_attribute}" | grep "linguist-generated=true" >/dev/null 2>/dev/null; then
-        pattern=$(echo ${git_attribute} | cut -d' ' -f1)
-        escaped_pattern=$(printf '%s\n' "$pattern" | sed -e 's/[\/&]/\\&/g')
-        # Delete known generated files
-        sed -i '/'"${escaped_pattern}"'/d' "${TMPFILE}"
-    fi
-done < "${GIT_TOP_DIR}/.gitattributes"
-
-echo "INFO: Showing non-generated files:"
-(
-    set -x
-    cat "${TMPFILE}"
-)
-
-# Get only files that have changed
-changed_files=$(cut -d' ' -f2 "${TMPFILE}" | xargs)
-
-details=$(git diff --shortstat "$ancestor" "${HEAD}" -- ${changed_files})
-add=$(echo "$details" | grep -o '[0-9]* insertion' | grep -o '[0-9]*' || true)
-remove=$(echo "$details" | grep -o '[0-9]* deletion' | grep -o '[0-9]*' || true)
-pr_size=0
-if [ "$add" ]; then
-  pr_size=$(("$pr_size" + "$add"))
-fi
-if [ "$remove" ]; then
-  pr_size=$(("$pr_size" + "$remove"))
-fi
-echo "INFO: PR SIZE is ${pr_size}"
-
-if ((pr_size > 2000)); then
-    echo
-    echo 'Your PR is '"$pr_size"' LOC which is more than the 2000 maximum'
-    echo 'allowed within PyTorch infra. PLease make sure to split up'
-    echo 'your PR into smaller pieces that can be reviewed.'
-    echo 'If you think that this rule should not apply to your PR,'
-    echo 'please contact @albanD or @seemethere.'
-    echo
-    exit 1
-fi

.github/scripts/process_commit.py

@@ -99,7 +99,7 @@ if __name__ == "__main__":
     repo_labels = get_repo_labels()
 
     primary_labels = set(filter(lambda x: x.startswith(PRIMARY_LABEL_FILTER), repo_labels))
-    has_both_labels = bool(primary_labels.intersection(labels)) and bool(SECONDARY_LABELS.intersection(labels))
+    has_both_labels = bool(primary_labels.intersection(labels) and SECONDARY_LABELS.intersection(labels))
     is_properly_labeled = has_both_labels or bool(ALLOWED_ONLY_SECONDARY.intersection(labels))
 
     if not is_properly_labeled:

.github/scripts/run_torchbench.py

@@ -5,25 +5,22 @@ Currently, only supports running tests on specified model names
 Testing environment:
 - Intel Xeon 8259CL @ 2.50 GHz, 24 Cores with disabled Turbo and HT
 - Nvidia Tesla T4
-- Nvidia Driver 470.82.01
-- Python 3.8
-- CUDA 11.3
+- Nvidia Driver 450.51.06
+- Python 3.7
+- CUDA 10.2
 """
 # Known issues:
 # 1. Does not reuse the build artifact in other CI workflows
 # 2. CI jobs are serialized because there is only one worker
 import os
-import boto3  # type: ignore[import]
 import git  # type: ignore[import]
 import pathlib
 import argparse
 import subprocess
-from pathlib import Path
 
-from typing import List, Tuple
+from typing import List
 
 TORCHBENCH_CONFIG_NAME = "config.yaml"
-TORCHBENCH_USERBENCHMARK_CONFIG_NAME = "ub-config.yaml"
 MAGIC_PREFIX = "RUN_TORCHBENCH:"
 MAGIC_TORCHBENCH_PREFIX = "TORCHBENCH_BRANCH:"
 ABTEST_CONFIG_TEMPLATE = """# This config is automatically generated by run_torchbench.py
@@ -33,25 +30,6 @@ threshold: 100
 direction: decrease
 timeout: 720
 tests:"""
-S3_BUCKET = "ossci-metrics"
-S3_PREFIX = "torchbench-pr-test"
-S3_URL_BASE = f"https://{S3_BUCKET}.s3.amazonaws.com/"
-
-class S3Client:
-    def __init__(self, bucket: str = S3_BUCKET, prefix: str = S3_PREFIX):
-        self.s3 = boto3.client('s3')
-        self.resource = boto3.resource('s3')
-        self.bucket = bucket
-        self.prefix = prefix
-
-    def upload_file(self, file_path: Path, filekey_prefix: str) -> None:
-        assert file_path.is_file(), f"Specified file path {file_path} does not exist or not file."
-        file_name = file_path.name
-        s3_key = f"{self.prefix}/{filekey_prefix}/{file_name}"
-        print(f"Uploading file {file_name} to S3 with key: {s3_key}")
-        self.s3.upload_file(str(file_path), self.bucket, s3_key)
-        # output the result URL
-        print(f"Uploaded the result file {file_name} to {S3_URL_BASE}{s3_key}")
 
 def gen_abtest_config(control: str, treatment: str, models: List[str]) -> str:
     d = {}
@@ -76,51 +54,33 @@ def find_current_branch(repo_path: str) -> str:
     name: str = repo.active_branch.name
     return name
 
-def deploy_torchbench_config(output_dir: str, config: str, config_name: str = TORCHBENCH_CONFIG_NAME) -> None:
+def deploy_torchbench_config(output_dir: str, config: str) -> None:
     # Create test dir if needed
     pathlib.Path(output_dir).mkdir(exist_ok=True)
     # TorchBench config file name
-    config_path = os.path.join(output_dir, config_name)
+    config_path = os.path.join(output_dir, TORCHBENCH_CONFIG_NAME)
     with open(config_path, "w") as fp:
         fp.write(config)
 
-def get_valid_models(torchbench_path: str) -> List[str]:
-    benchmark_path = os.path.join(torchbench_path, "torchbenchmark", "models")
-    valid_models = [model for model in os.listdir(benchmark_path) if os.path.isdir(os.path.join(benchmark_path, model))]
-    return valid_models
-
-def get_valid_userbenchmarks(torchbench_path: str) -> List[str]:
-    def is_valid_ub_dir(ub_path: str) -> bool:
-        return os.path.isdir(ub_path) and os.path.exists(os.path.join(ub_path, "__init__.py"))
-    ub_path = os.path.join(os.path.abspath(torchbench_path), "userbenchmark")
-    ubs = list(filter(is_valid_ub_dir, [os.path.join(ub_path, ubdir) for ubdir in os.listdir(ub_path)]))
-    valid_ubs = list(map(lambda x: os.path.basename(x), ubs))
-    return valid_ubs
-
-def extract_models_from_pr(torchbench_path: str, prbody_file: str) -> Tuple[List[str], List[str]]:
+def extract_models_from_pr(torchbench_path: str, prbody_file: str) -> List[str]:
     model_list = []
-    userbenchmark_list = []
-    pr_list = []
     with open(prbody_file, "r") as pf:
         lines = map(lambda x: x.strip(), pf.read().splitlines())
         magic_lines = list(filter(lambda x: x.startswith(MAGIC_PREFIX), lines))
         if magic_lines:
             # Only the first magic line will be recognized.
-            pr_list = list(map(lambda x: x.strip(), magic_lines[0][len(MAGIC_PREFIX):].split(",")))
-    valid_models = get_valid_models(torchbench_path)
-    valid_ubs = get_valid_userbenchmarks(torchbench_path)
-    for pr_bm in pr_list:
-        if pr_bm in valid_models or pr_bm == "ALL":
-            model_list.append(pr_bm)
-        elif pr_bm in valid_ubs:
-            userbenchmark_list.append(pr_bm)
-        else:
-            print(f"The model or benchmark {pr_bm} you specified does not exist in TorchBench suite. Please double check.")
-            exit(-1)
-    # Shortcut: if pr_list is ["ALL"], run all the model tests
-    if "ALL" in model_list:
-        model_list = ["ALL"]
-    return model_list, userbenchmark_list
+            model_list = list(map(lambda x: x.strip(), magic_lines[0][len(MAGIC_PREFIX):].split(",")))
+    # Shortcut: if model_list is ["ALL"], run all the tests
+    if model_list == ["ALL"]:
+        return model_list
+    # Sanity check: make sure all the user specified models exist in torchbench repository
+    benchmark_path = os.path.join(torchbench_path, "torchbenchmark", "models")
+    full_model_list = [model for model in os.listdir(benchmark_path) if os.path.isdir(os.path.join(benchmark_path, model))]
+    for m in model_list:
+        if m not in full_model_list:
+            print(f"The model {m} you specified does not exist in TorchBench suite. Please double check.")
+            return []
+    return model_list
 
 def find_torchbench_branch(prbody_file: str) -> str:
     branch_name: str = ""
@@ -140,39 +100,13 @@ def run_torchbench(pytorch_path: str, torchbench_path: str, output_dir: str) ->
     env = dict(os.environ)
     command = ["python", "bisection.py", "--work-dir", output_dir,
                "--pytorch-src", pytorch_path, "--torchbench-src", torchbench_path,
-               "--config", os.path.join(output_dir, TORCHBENCH_CONFIG_NAME),
+               "--config", os.path.join(output_dir, "config.yaml"),
                "--output", os.path.join(output_dir, "result.txt")]
-    print(f"Running torchbench command: {command}")
     subprocess.check_call(command, cwd=torchbench_path, env=env)
 
-def run_userbenchmarks(pytorch_path: str, torchbench_path: str, base_sha: str, head_sha: str,
-                       userbenchmark: str, output_dir: str) -> None:
-    # Copy system environment so that we will not override
-    env = dict(os.environ)
-    command = ["python", "./.github/scripts/abtest.py",
-               "--pytorch-repo", pytorch_path,
-               "--base", base_sha,
-               "--head", head_sha,
-               "--userbenchmark", userbenchmark,
-               "--output-dir", output_dir]
-    print(f"Running torchbench userbenchmark command: {command}")
-    subprocess.check_call(command, cwd=torchbench_path, env=env)
-
-def process_upload_s3(result_dir: str) -> None:
-    # validate result directory
-    result_dir_path = Path(result_dir)
-    assert result_dir_path.exists(), f"Specified result directory {result_dir} doesn't exist."
-    # upload all files to S3 bucket oss-ci-metrics
-    files = [x for x in result_dir_path.iterdir() if x.is_file()]
-    # upload file to S3 bucket
-    s3_client: S3Client = S3Client()
-    filekey_prefix = result_dir_path.name
-    for f in files:
-        s3_client.upload_file(f, filekey_prefix)
-
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description='Run TorchBench tests based on PR')
-    parser.add_argument('--pr-body', help="The file that contains body of a Pull Request")
+    parser.add_argument('--pr-body', required=True, help="The file that contains body of a Pull Request")
 
     subparsers = parser.add_subparsers(dest='command')
     # parser for setup the torchbench branch name env
@@ -184,9 +118,6 @@ if __name__ == "__main__":
     run_parser.add_argument('--pr-head-sha', required=True, type=str, help="The Pull Request head hash")
     run_parser.add_argument('--pytorch-path', required=True, type=str, help="Path to pytorch repository")
     run_parser.add_argument('--torchbench-path', required=True, type=str, help="Path to TorchBench repository")
-    # parser to upload results to S3
-    upload_parser = subparsers.add_parser("upload-s3")
-    upload_parser.add_argument('--result-dir', required=True, type=str, help="Path to benchmark output")
     args = parser.parse_args()
 
     if args.command == 'set-torchbench-branch':
@@ -195,30 +126,21 @@ if __name__ == "__main__":
         setup_gha_env(MAGIC_TORCHBENCH_PREFIX[:-1], branch_name)
     elif args.command == 'run':
         output_dir: str = os.path.join(os.environ["HOME"], ".torchbench", "bisection", f"pr{args.pr_num}")
+        # Identify the specified models and verify the input
+        models = extract_models_from_pr(args.torchbench_path, args.pr_body)
+        if not models:
+            print("Can't parse the model filter from the pr body. Currently we only support allow-list.")
+            exit(-1)
         # Assert the current branch in args.torchbench_path is the same as the one specified in pr body
         branch_name = find_torchbench_branch(args.pr_body)
         current_branch = find_current_branch(args.torchbench_path)
         assert branch_name == current_branch, f"Torchbench repo {args.torchbench_path} is on branch {current_branch}, \
                                                 but user specified to run on branch {branch_name}."
         print(f"Ready to run TorchBench with benchmark. Result will be saved in the directory: {output_dir}.")
-        # Identify the specified models and userbenchmarks
-        models, userbenchmarks = extract_models_from_pr(args.torchbench_path, args.pr_body)
-        if models:
-            torchbench_config = gen_abtest_config(args.pr_base_sha, args.pr_head_sha, models)
-            deploy_torchbench_config(output_dir, torchbench_config)
-            run_torchbench(pytorch_path=args.pytorch_path, torchbench_path=args.torchbench_path, output_dir=output_dir)
-        if userbenchmarks:
-            assert len(userbenchmarks) == 1, \
-                "We don't support running multiple userbenchmarks in single workflow yet." \
-                "If you need, please submit a feature request."
-            run_userbenchmarks(pytorch_path=args.pytorch_path, torchbench_path=args.torchbench_path,
-                               base_sha=args.pr_base_sha, head_sha=args.pr_head_sha,
-                               userbenchmark=userbenchmarks[0], output_dir=output_dir)
-        if not models and not userbenchmarks:
-            print("Can't parse valid models or userbenchmarks from the pr body. Quit.")
-            exit(-1)
-    elif args.command == 'upload-s3':
-        process_upload_s3(args.result_dir)
+        # Run TorchBench with the generated config
+        torchbench_config = gen_abtest_config(args.pr_base_sha, args.pr_head_sha, models)
+        deploy_torchbench_config(output_dir, torchbench_config)
+        run_torchbench(pytorch_path=args.pytorch_path, torchbench_path=args.torchbench_path, output_dir=output_dir)
     else:
         print(f"The command {args.command} is not supported.")
         exit(-1)

.github/scripts/syncbranches.py

@@ -0,0 +1,25 @@
+#!/usr/bin/env python3
+
+from gitutils import get_git_repo_dir, get_git_remote_name, GitRepo
+from typing import Any
+
+
+def parse_args() -> Any:
+    from argparse import ArgumentParser
+    parser = ArgumentParser("Merge PR/branch into default branch")
+    parser.add_argument("--sync-branch", default="sync")
+    parser.add_argument("--default-branch", type=str, default="main")
+    parser.add_argument("--dry-run", action="store_true")
+    parser.add_argument("--debug", action="store_true")
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    repo = GitRepo(get_git_repo_dir(), get_git_remote_name(), debug=args.debug)
+    repo.cherry_pick_commits(args.sync_branch, args.default_branch)
+    repo.push(args.default_branch, args.dry_run)
+
+
+if __name__ == '__main__':
+    main()

.github/scripts/test_fetch_latest_green_commit.py

@@ -1,101 +0,0 @@
-from unittest import TestCase, main, mock
-from typing import Any, List, Dict
-from fetch_latest_green_commit import isGreen, WorkflowCheck
-
-workflowNames = [
-    "pull",
-    "trunk",
-    "Lint",
-    "linux-binary-libtorch-pre-cxx11",
-    "android-tests",
-    "windows-binary-wheel",
-    "periodic",
-    "docker-release-builds",
-    "nightly",
-    "pr-labels",
-    "Close stale pull requests",
-    "Update S3 HTML indices for download.pytorch.org",
-    "Create Release"
-]
-
-def set_workflow_job_status(workflow: List[Dict[str, Any]], name: str, status: str) -> List[Dict[str, Any]]:
-    for check in workflow:
-        if check['workflowName'] == name:
-            check['conclusion'] = status
-    return workflow
-
-class TestChecks:
-    def make_test_checks(self) -> List[Dict[str, Any]]:
-        workflow_checks = []
-        for i in range(len(workflowNames)):
-            workflow_checks.append(WorkflowCheck(
-                workflowName=workflowNames[i],
-                name="test/job",
-                jobName="job",
-                conclusion="success",
-            )._asdict())
-        return workflow_checks
-
-class TestPrintCommits(TestCase):
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_all_successful(self, mock_get_commit_results: Any) -> None:
-        "Test with workflows are successful"
-        workflow_checks = mock_get_commit_results()
-        self.assertTrue(isGreen("sha", workflow_checks)[0])
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_necessary_successful(self, mock_get_commit_results: Any) -> None:
-        "Test with necessary workflows are successful"
-        workflow_checks = mock_get_commit_results()
-        workflow_checks = set_workflow_job_status(workflow_checks, workflowNames[8], "failed")
-        workflow_checks = set_workflow_job_status(workflow_checks, workflowNames[9], "failed")
-        workflow_checks = set_workflow_job_status(workflow_checks, workflowNames[10], "failed")
-        workflow_checks = set_workflow_job_status(workflow_checks, workflowNames[11], "failed")
-        workflow_checks = set_workflow_job_status(workflow_checks, workflowNames[12], "failed")
-        self.assertTrue(isGreen("sha", workflow_checks)[0])
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_necessary_skipped(self, mock_get_commit_results: Any) -> None:
-        "Test with necessary job (ex: pull) skipped"
-        workflow_checks = mock_get_commit_results()
-        workflow_checks = set_workflow_job_status(workflow_checks, "pull", "skipped")
-        result = isGreen("sha", workflow_checks)
-        self.assertTrue(result[0])
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_skippable_skipped(self, mock_get_commit_results: Any) -> None:
-        "Test with skippable jobs (periodic and docker-release-builds skipped"
-        workflow_checks = mock_get_commit_results()
-        workflow_checks = set_workflow_job_status(workflow_checks, "periodic", "skipped")
-        workflow_checks = set_workflow_job_status(workflow_checks, "docker-release-builds", "skipped")
-        self.assertTrue(isGreen("sha", workflow_checks))
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_necessary_failed(self, mock_get_commit_results: Any) -> None:
-        "Test with necessary job (ex: Lint) failed"
-        workflow_checks = mock_get_commit_results()
-        workflow_checks = set_workflow_job_status(workflow_checks, "Lint", "failed")
-        result = isGreen("sha", workflow_checks)
-        self.assertFalse(result[0])
-        self.assertEqual(result[1], "Lint checks were not successful")
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value=TestChecks().make_test_checks())
-    def test_skippable_failed(self, mock_get_commit_results: Any) -> None:
-        "Test with skippable job (ex: docker-release-builds) failing"
-        workflow_checks = mock_get_commit_results()
-        workflow_checks = set_workflow_job_status(workflow_checks, "periodic", "skipped")
-        workflow_checks = set_workflow_job_status(workflow_checks, "docker-release-builds", "failed")
-        result = isGreen("sha", workflow_checks)
-        self.assertFalse(result[0])
-        self.assertEqual(result[1], "docker-release-builds checks were not successful")
-
-    @mock.patch('fetch_latest_green_commit.get_commit_results', return_value={})
-    def test_no_workflows(self, mock_get_commit_results: Any) -> None:
-        "Test with missing workflows"
-        workflow_checks = mock_get_commit_results()
-        result = isGreen("sha", workflow_checks)
-        self.assertFalse(result[0])
-        self.assertEqual(result[1], "missing required workflows: pull, trunk, lint, linux-binary, windows-binary")
-
-if __name__ == "__main__":
-    main()

.github/scripts/test_filter_test_configs.py

@@ -1,88 +0,0 @@
-#!/usr/bin/env python3
-
-import os
-import yaml
-import json
-from unittest import TestCase, main, mock
-from filter_test_configs import get_labels, filter, PREFIX, VALID_TEST_CONFIG_LABELS
-import requests
-from requests.models import Response
-from typing import Any, Dict
-
-
-def mocked_gh_get_labels_failed(url: str, headers: Dict[str, str]) -> Response:
-    mocked_response = Response()
-    mocked_response.status_code = requests.codes.bad_request
-    return mocked_response
-
-
-def mocked_gh_get_labels(url: str, headers: Dict[str, str]) -> Response:
-    mocked_response = Response()
-    mocked_response.status_code = requests.codes.ok
-    mocked_response._content = b'[{"name": "foo"}, {"name": "bar"}, {}, {"name": ""}]'
-    return mocked_response
-
-
-class TestConfigFilter(TestCase):
-
-    def setUp(self) -> None:
-        os.environ["GITHUB_TOKEN"] = "GITHUB_TOKEN"
-
-    @mock.patch("filter_test_configs.requests.get", side_effect=mocked_gh_get_labels)
-    def test_get_labels(self, mocked_gh: Any) -> None:
-        labels = get_labels(pr_number=12345)
-        self.assertSetEqual({"foo", "bar"}, labels)
-
-    @mock.patch("filter_test_configs.requests.get", side_effect=mocked_gh_get_labels_failed)
-    def test_get_labels_failed(self, mocked_gh: Any) -> None:
-        labels = get_labels(pr_number=54321)
-        self.assertFalse(labels)
-
-    def test_filter(self) -> None:
-        mocked_labels = {f"{PREFIX}cfg", "ciflow/trunk", "plain-cfg"}
-        testcases = [
-            {
-                "test_matrix": '{include: [{config: "default", runner: "linux"}]}',
-                "expected": '{"include": [{"config": "default", "runner": "linux"}]}',
-                "description": "No match, keep the same test matrix",
-            },
-            {
-                "test_matrix": '{include: [{config: "default", runner: "linux"}, {config: "plain-cfg"}]}',
-                "expected": '{"include": [{"config": "default", "runner": "linux"}, {"config": "plain-cfg"}]}',
-                "description": "No match because there is no prefix or suffix, keep the same test matrix",
-            },
-            {
-                "test_matrix": '{include: [{config: "default", runner: "linux"}, {config: "cfg", shard: 1}]}',
-                "expected": '{"include": [{"config": "cfg", "shard": 1}]}',
-                "description": "Found a match, only keep that",
-            },
-        ]
-
-        for case in testcases:
-            filtered_test_matrix = filter(yaml.safe_load(case["test_matrix"]), mocked_labels)
-            self.assertEqual(case["expected"], json.dumps(filtered_test_matrix))
-
-    def test_filter_with_valid_label(self) -> None:
-        mocked_labels = {f"{PREFIX}cfg", "ciflow/trunk"}
-        VALID_TEST_CONFIG_LABELS.add(f"{PREFIX}cfg")
-
-        testcases = [
-            {
-                "test_matrix": '{include: [{config: "default", runner: "linux"}]}',
-                "expected": '{"include": []}',
-                "description": "Found a valid label in the PR body, return the filtered test matrix",
-            },
-            {
-                "test_matrix": '{include: [{config: "default", runner: "linux"}, {config: "cfg", shard: 1}]}',
-                "expected": '{"include": [{"config": "cfg", "shard": 1}]}',
-                "description": "Found a match, only keep that",
-            },
-        ]
-
-        for case in testcases:
-            filtered_test_matrix = filter(yaml.safe_load(case["test_matrix"]), mocked_labels)
-            self.assertEqual(case["expected"], json.dumps(filtered_test_matrix))
-
-
-if __name__ == '__main__':
-    main()

.github/scripts/test_gitutils.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-from gitutils import PeekableIterator, patterns_to_regex
+from gitutils import PeekableIterator
 from unittest import TestCase, main
 
 class TestPeekableIterator(TestCase):
@@ -22,18 +22,6 @@ class TestPeekableIterator(TestCase):
                 self.assertTrue(iter_.peek() is None)
 
 
-class TestPattern(TestCase):
-    def test_double_asterisks(self) -> None:
-        allowed_patterns = [
-            "aten/src/ATen/native/**LinearAlgebra*",
-        ]
-        patterns_re = patterns_to_regex(allowed_patterns)
-        fnames = [
-            "aten/src/ATen/native/LinearAlgebra.cpp",
-            "aten/src/ATen/native/cpu/LinearAlgebraKernel.cpp"]
-        for filename in fnames:
-            self.assertTrue(patterns_re.match(filename))
-
 
 if __name__ == '__main__':
     main()

.github/scripts/test_trymerge.py

@@ -11,28 +11,12 @@ import json
 import os
 from hashlib import sha256
 
-from trymerge import (find_matching_merge_rule,
-                      get_land_checkrun_conclusions,
-                      validate_land_time_checks,
-                      gh_graphql,
-                      gh_get_team_members,
-                      read_merge_rules,
-                      validate_revert,
-                      filter_pending_checks,
-                      filter_failed_checks,
-                      GitHubPR,
-                      MergeRule,
-                      MandatoryChecksMissingError,
-                      WorkflowCheckState,
-                      main as trymerge_main)
+from trymerge import find_matching_merge_rule, gh_graphql, gh_get_team_members, GitHubPR, MergeRule, MandatoryChecksMissingError
 from gitutils import get_git_remote_name, get_git_repo_dir, GitRepo
-from typing import Any, List, Optional
+from typing import cast, Any, List, Optional
 from unittest import TestCase, main, mock
 from urllib.error import HTTPError
 
-if 'GIT_REMOTE_URL' not in os.environ:
-    os.environ['GIT_REMOTE_URL'] = "https://github.com/pytorch/pytorch"
-
 def mocked_gh_graphql(query: str, **kwargs: Any) -> Any:
     gql_db_fname = os.path.join(os.path.dirname(__file__), "gql_mocks.json")
 
@@ -69,110 +53,39 @@ def mocked_gh_graphql(query: str, **kwargs: Any) -> Any:
 
     return rc
 
-def mock_parse_args(revert: bool = False,
-                    force: bool = False) -> Any:
-    class Object(object):
-        def __init__(self) -> None:
-            self.revert = revert
-            self.force = force
-            self.pr_num = 76123
-            self.dry_run = True
-            self.comment_id = 0
-            self.on_mandatory = False
-            self.on_green = False
-            self.land_checks = False
-            self.reason = 'this is for testing'
 
-    return Object()
-
-def mock_revert(repo: GitRepo, pr: GitHubPR, *,
-                dry_run: bool = False,
-                comment_id: Optional[int] = None,
-                reason: Optional[str] = None) -> None:
-    pass
-
-def mock_merge(pr_num: int, repo: GitRepo,
-               dry_run: bool = False,
-               skip_mandatory_checks: bool = False,
-               comment_id: Optional[int] = None,
-               mandatory_only: bool = False,
-               on_green: bool = False,
-               land_checks: bool = False,
-               timeout_minutes: int = 400,
-               stale_pr_days: int = 3) -> None:
-    pass
-
-def mock_gh_get_info() -> Any:
-    return {"closed": False, "isCrossRepository": False}
-
-
-def mocked_read_merge_rules_NE(repo: Any, org: str, project: str) -> List[MergeRule]:
-    return [
-        MergeRule(name="mock with nonexistent check",
-                  patterns=["*"],
-                  approved_by=[],
-                  mandatory_checks_name=["Lint",
-                                         "Facebook CLA Check",
-                                         "nonexistent"],
-                  ),
+def mocked_read_merge_rules(repo: Optional[GitRepo], org: str, project: str) -> List[MergeRule]:
+    mock_merge_rules = """
+    [
+        {
+            "name": "mock with nonexistent check",
+            "patterns": ["*"],
+            "approved_by": [],
+            "mandatory_checks_name": [
+                "Facebook CLA Check",
+                "Lint",
+                "nonexistent"
+            ]
+        }
     ]
+    """
+    rc = json.loads(mock_merge_rules, object_hook=lambda x: MergeRule(**x))
+    return cast(List[MergeRule], rc)
 
 
-def mocked_read_merge_rules(repo: Any, org: str, project: str) -> List[MergeRule]:
-    return [
-        MergeRule(name="super",
-                  patterns=["*"],
-                  approved_by=["pytorch/metamates"],
-                  mandatory_checks_name=["Lint",
-                                         "Facebook CLA Check",
-                                         "pull / linux-xenial-cuda11.3-py3.7-gcc7 / build",
-                                         ],
-                  ),
-    ]
-
-
-def mocked_read_merge_rules_raise(repo: Any, org: str, project: str) -> List[MergeRule]:
-    raise RuntimeError("testing")
-
-
-class DummyGitRepo(GitRepo):
-    def __init__(self) -> None:
-        super().__init__(get_git_repo_dir(), get_git_remote_name())
-
-    def commits_resolving_gh_pr(self, pr_num: int) -> List[str]:
-        return ["FakeCommitSha"]
-
-    def commit_message(self, ref: str) -> str:
-        return "super awsome commit message"
-
 class TestGitHubPR(TestCase):
-    def test_merge_rules_valid(self) -> None:
-        "Test that merge_rules.yaml can be parsed"
-        repo = DummyGitRepo()
-        self.assertGreater(len(read_merge_rules(repo, "pytorch", "pytorch")), 1)
-
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules)
-    def test_match_rules(self, mocked_gql: Any, mocked_rmr: Any) -> None:
+    def test_match_rules(self, mocked_gql: Any) -> None:
         "Tests that PR passes merge rules"
-        pr = GitHubPR("pytorch", "pytorch", 77700)
-        repo = DummyGitRepo()
+        pr = GitHubPR("pytorch", "pytorch", 71759)
+        repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
         self.assertTrue(find_matching_merge_rule(pr, repo) is not None)
 
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules_raise)
-    def test_read_merge_rules_fails(self, mocked_gql: Any, mocked_rmr: Any) -> None:
-        "Tests that PR fails to read the merge rules"
-        pr = GitHubPR("pytorch", "pytorch", 77700)
-        repo = DummyGitRepo()
-        self.assertRaisesRegex(RuntimeError, "testing", lambda: find_matching_merge_rule(pr, repo))
-
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules)
-    def test_lint_fails(self, mocked_gql: Any, mocked_rmr: Any) -> None:
+    def test_lint_fails(self, mocked_gql: Any) -> None:
         "Tests that PR fails mandatory lint check"
         pr = GitHubPR("pytorch", "pytorch", 74649)
-        repo = DummyGitRepo()
+        repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
         self.assertRaises(RuntimeError, lambda: find_matching_merge_rule(pr, repo))
 
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
@@ -219,7 +132,7 @@ class TestGitHubPR(TestCase):
     def test_checksuites_pagination(self, mocked_gql: Any) -> None:
         "Tests that PR with lots of checksuits can be fetched"
         pr = GitHubPR("pytorch", "pytorch", 73811)
-        self.assertEqual(len(pr.get_checkrun_conclusions()), 107)
+        self.assertGreater(len(pr.get_checkrun_conclusions()), 0)
 
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
     def test_comments_pagination(self, mocked_gql: Any) -> None:
@@ -256,13 +169,13 @@ class TestGitHubPR(TestCase):
         self.assertGreater(len(authors), 50)
         self.assertTrue("@" in pr.get_author())
 
-    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules_NE)
+    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules)
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
     def test_pending_status_check(self, mocked_gql: Any, mocked_read_merge_rules: Any) -> None:
         """ Tests that PR with nonexistent/pending status checks fails with the right reason.
         """
         pr = GitHubPR("pytorch", "pytorch", 76118)
-        repo = DummyGitRepo()
+        repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
         self.assertRaisesRegex(MandatoryChecksMissingError,
                                ".*are pending/not yet run.*",
                                lambda: find_matching_merge_rule(pr, repo))
@@ -283,92 +196,8 @@ class TestGitHubPR(TestCase):
         """
         pr = GitHubPR("pytorch", "pytorch", 77700)
         conclusions = pr.get_checkrun_conclusions()
-        self.assertEqual(len(conclusions), 83)
-        self.assertTrue("pull / linux-docs / build-docs (cpp)" in conclusions.keys())
+        self.assertTrue("linux-docs / build-docs (cpp)" in conclusions.keys())
 
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    def test_cancelled_gets_ignored(self, mocked_gql: Any) -> None:
-        """ Tests that cancelled workflow does not override existing successfull status
-        """
-        pr = GitHubPR("pytorch", "pytorch", 82169)
-        conclusions = pr.get_checkrun_conclusions()
-        self.assertTrue("Lint" in conclusions.keys())
-        self.assertEqual(conclusions["Lint"][0], "SUCCESS")
-
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    def test_get_many_land_checks(self, mocked_gql: Any) -> None:
-        """ Tests that all checkruns can be fetched for a commit
-        """
-        conclusions = get_land_checkrun_conclusions('pytorch', 'pytorch', '6882717f73deffb692219ccd1fd6db258d8ed684')
-        self.assertEqual(len(conclusions), 101)
-        self.assertTrue("pull / linux-docs / build-docs (cpp)" in conclusions.keys())
-
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    def test_failed_land_checks(self, mocked_gql: Any) -> None:
-        """ Tests that PR with Land Checks fail with a RunTime error
-        """
-        self.assertRaisesRegex(RuntimeError,
-                               ".*Failed to merge; some land checks failed.*",
-                               lambda: validate_land_time_checks('pytorch', 'pytorch', '6882717f73deffb692219ccd1fd6db258d8ed684'))
-
-    @mock.patch('trymerge.gh_get_pr_info', return_value=mock_gh_get_info())
-    @mock.patch('trymerge.parse_args', return_value=mock_parse_args(True, False))
-    @mock.patch('trymerge.try_revert', side_effect=mock_revert)
-    def test_main_revert(self, mock_revert: Any, mock_parse_args: Any, gh_get_pr_info: Any) -> None:
-        trymerge_main()
-        mock_revert.assert_called_once()
-
-    @mock.patch('trymerge.gh_get_pr_info', return_value=mock_gh_get_info())
-    @mock.patch('trymerge.parse_args', return_value=mock_parse_args(False, True))
-    @mock.patch('trymerge.merge', side_effect=mock_merge)
-    def test_main_force(self, mock_merge: Any, mock_parse_args: Any, mock_gh_get_info: Any) -> None:
-        trymerge_main()
-        mock_merge.assert_called_once_with(mock.ANY,
-                                           mock.ANY,
-                                           dry_run=mock.ANY,
-                                           skip_mandatory_checks=True,
-                                           comment_id=mock.ANY,
-                                           on_green=False,
-                                           land_checks=False,
-                                           mandatory_only=False)
-
-    @mock.patch('trymerge.gh_get_pr_info', return_value=mock_gh_get_info())
-    @mock.patch('trymerge.parse_args', return_value=mock_parse_args(False, False))
-    @mock.patch('trymerge.merge', side_effect=mock_merge)
-    def test_main_merge(self, mock_merge: Any, mock_parse_args: Any, mock_gh_get_info: Any) -> None:
-        trymerge_main()
-        mock_merge.assert_called_once_with(mock.ANY,
-                                           mock.ANY,
-                                           dry_run=mock.ANY,
-                                           skip_mandatory_checks=False,
-                                           comment_id=mock.ANY,
-                                           on_green=False,
-                                           land_checks=False,
-                                           mandatory_only=False)
-
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    @mock.patch('trymerge.read_merge_rules', side_effect=mocked_read_merge_rules)
-    def test_revert_rules(self, mock_gql: Any, mock_mr: Any) -> None:
-        """ Tests that reverts from collaborators are allowed """
-        pr = GitHubPR("pytorch", "pytorch", 79694)
-        repo = DummyGitRepo()
-        self.assertIsNotNone(validate_revert(repo, pr, comment_id=1189459845))
-
-    def test_checks_filter(self) -> None:
-        checks = [
-            WorkflowCheckState(name="check0", status="SUCCESS", url="url0"),
-            WorkflowCheckState(name="check1", status="FAILURE", url="url1"),
-            WorkflowCheckState(name="check2", status="STARTUP_FAILURE", url="url2"),
-            WorkflowCheckState(name="check3", status=None, url="url3"),
-        ]
-
-        checks_dict = {check.name : check for check in checks}
-
-        pending_checks = filter_pending_checks(checks_dict)
-        failing_checks = filter_failed_checks(checks_dict)
-
-        self.assertListEqual(failing_checks, [checks[1], checks[2]])
-        self.assertListEqual(pending_checks, [checks[3]])
 
 if __name__ == "__main__":
     main()

.github/scripts/test_tryrebase.py

@@ -15,28 +15,12 @@ class TestRebase(TestCase):
         "Tests rebase successfully"
         pr = GitHubPR("pytorch", "pytorch", 31093)
         repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
-        rebase_onto(pr, repo, 'master')
+        rebase_onto(pr, repo)
         calls = [mock.call('fetch', 'origin', 'pull/31093/head:pull/31093/head'),
-                 mock.call('rebase', 'refs/remotes/origin/master', 'pull/31093/head'),
+                 mock.call('rebase', 'master', 'pull/31093/head'),
                  mock.call('push', '-f', 'https://github.com/mingxiaoh/pytorch.git', 'pull/31093/head:master')]
         mocked_run_git.assert_has_calls(calls)
-        self.assertTrue(
-            "Successfully rebased `master` onto `refs/remotes/origin/master`" in mocked_post_comment.call_args[0][3])
-
-    @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
-    @mock.patch('gitutils.GitRepo._run_git')
-    @mock.patch('tryrebase.gh_post_comment')
-    def test_rebase_to_stable(self, mocked_post_comment: Any, mocked_run_git: Any, mocked_gql: Any) -> None:
-        "Tests rebase to viable/strict successfully"
-        pr = GitHubPR("pytorch", "pytorch", 31093)
-        repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
-        rebase_onto(pr, repo, 'viable/strict', False)
-        calls = [mock.call('fetch', 'origin', 'pull/31093/head:pull/31093/head'),
-                 mock.call('rebase', 'refs/remotes/origin/viable/strict', 'pull/31093/head'),
-                 mock.call('push', '-f', 'https://github.com/mingxiaoh/pytorch.git', 'pull/31093/head:master')]
-        mocked_run_git.assert_has_calls(calls)
-        self.assertTrue(
-            "Successfully rebased `master` onto `refs/remotes/origin/viable/strict`" in mocked_post_comment.call_args[0][3])
+        self.assertTrue("Successfully rebased `master` onto `master`" in mocked_post_comment.call_args[0][3])
 
     @mock.patch('trymerge.gh_graphql', side_effect=mocked_gh_graphql)
     @mock.patch('gitutils.GitRepo._run_git', return_value="Everything up-to-date")
@@ -45,9 +29,9 @@ class TestRebase(TestCase):
         "Tests branch already up to date"
         pr = GitHubPR("pytorch", "pytorch", 31093)
         repo = GitRepo(get_git_repo_dir(), get_git_remote_name())
-        rebase_onto(pr, repo, 'master')
+        rebase_onto(pr, repo)
         calls = [mock.call('fetch', 'origin', 'pull/31093/head:pull/31093/head'),
-                 mock.call('rebase', 'refs/remotes/origin/master', 'pull/31093/head'),
+                 mock.call('rebase', 'master', 'pull/31093/head'),
                  mock.call('push', '-f', 'https://github.com/mingxiaoh/pytorch.git', 'pull/31093/head:master')]
         mocked_run_git.assert_has_calls(calls)
         self.assertTrue(

.github/scripts/trymerge_explainer.py

@@ -1,146 +0,0 @@
-import os
-import re
-from typing import List, Pattern, Tuple, Optional
-
-
-BOT_COMMANDS_WIKI = "https://github.com/pytorch/pytorch/wiki/Bot-commands"
-
-CIFLOW_LABEL = re.compile(r"^ciflow/.+")
-CIFLOW_TRUNK_LABEL = re.compile(r"^ciflow/trunk")
-
-OFFICE_HOURS_LINK = "https://github.com/pytorch/pytorch/wiki/Dev-Infra-Office-Hours"
-CONTACT_US = f"Please reach out to the [PyTorch DevX Team]({OFFICE_HOURS_LINK}) with feedback or questions!"
-ALTERNATIVES = (
-    "If this is not the intended behavior, feel free to use some "
-    + f"of the other merge options in the [wiki]({BOT_COMMANDS_WIKI})."
-)
-LAND_CHECK_ROLLOUT = "https://github.com/pytorch/test-infra/blob/main/torchci/lib/bot/rolloutUtils.ts#L1-L34"
-
-
-def has_label(labels: List[str], pattern: Pattern[str] = CIFLOW_LABEL) -> bool:
-    return len(list(filter(pattern.match, labels))) > 0
-
-
-class TryMergeExplainer(object):
-    force: bool
-    on_green: bool
-    land_checks: bool
-    labels: List[str]
-    pr_num: int
-    org: str
-    project: str
-
-    has_trunk_label: bool
-    has_ciflow_label: bool
-
-    def __init__(
-        self,
-        force: bool,
-        on_green: bool,
-        land_checks: bool,
-        labels: List[str],
-        pr_num: int,
-        org: str,
-        project: str,
-    ):
-        self.force = force
-        self.on_green = on_green
-        self.land_checks = land_checks
-        self.labels = labels
-        self.pr_num = pr_num
-        self.org = org
-        self.project = project
-        self.get_flags()
-
-    def get_flags(self) -> Tuple[bool, bool]:
-        self.has_trunk_label = has_label(self.labels, CIFLOW_TRUNK_LABEL)
-        self.has_ciflow_label = has_label(self.labels, CIFLOW_LABEL)
-        should_check_land_branch = self.land_checks and not self.has_trunk_label
-        should_check_green = self.on_green or self.has_ciflow_label
-
-        return (should_check_green, should_check_land_branch)
-
-    def _get_flag_msg(self) -> str:
-        if self.force:
-            return " the force (-f) flag."
-        elif self.on_green:
-            return " the green (-g) flag."
-        elif self.land_checks:
-            return (
-                " the land checks (-l) flag."
-                + " If you did not specify this flag yourself, "
-                + f" you are likely enrolled in the [land checks rollout]({LAND_CHECK_ROLLOUT})."
-            )
-        else:
-            return "out a flag."
-
-    def _get_land_check_progress(self, commit: Optional[str]) -> str:
-        if commit is not None:
-            return (
-                " and land check "
-                + f"progress [here](https://hud.pytorch.org/{self.org}/{self.project}/commit/{commit})"
-            )
-        else:
-            return ""
-
-    def _get_flag_explanation_message(self) -> str:
-        if self.force:
-            return "This means your change will be merged **immediately**, bypassing any CI checks (ETA: 1-5 minutes)."
-        elif self.on_green:
-            return "This means that your change will be merged once all checks on your PR have passed (ETA: 0-4 Hours)."
-        elif self.land_checks:
-            if self.has_trunk_label:
-                land_check_msg_suffix = "have passed since you have added the `ciflow/trunk` label to your PR (ETA 0-4 Hours)."
-            else:
-                land_check_msg_suffix = (
-                    "and the land checks have passed (**ETA 4 Hours**). "
-                )
-                land_check_msg_suffix += "If you need to coordinate lands between different changes and cannot risk a land race, "
-                land_check_msg_suffix += "please add the `ciflow/trunk` label to your PR and wait for signal to complete, "
-                land_check_msg_suffix += "and then land your changes in proper order."
-                land_check_msg_suffix += (
-                    " Having `trunk`, `pull`, and `Lint` pre-run on a "
-                )
-                land_check_msg_suffix += (
-                    "PR will bypass land checks and the ETA should be immediate."
-                )
-
-            return (
-                "This means that your change will be merged once all checks on your PR "
-                + land_check_msg_suffix
-            )
-        else:
-            return "This means that your change will be merged once all checks on your PR have passed (ETA: 0-4 Hours)."
-
-    def get_merge_message(self, commit: Optional[str] = None) -> str:
-        message_prefix = "@pytorchbot successfully started a merge job."
-        progress_links = f"Check the current status [here]({os.getenv('GH_RUN_URL')}){self._get_land_check_progress(commit)}."
-        flag_message = f"The merge job was triggered with{self._get_flag_msg()}"
-        explanation_message = self._get_flag_explanation_message()
-
-        msg = message_prefix + " "
-        msg += progress_links + "\n"
-        msg += flag_message + " "
-        msg += explanation_message + " "
-        msg += ALTERNATIVES + "\n"
-        msg += CONTACT_US
-        return msg
-
-
-def get_revert_message(org: str, project: str, pr_num: int) -> str:
-    msg = (
-        "@pytorchbot successfully started a revert job."
-        + f" Check the current status [here]({os.getenv('GH_RUN_URL')}).\n"
-    )
-    msg += CONTACT_US
-    return msg
-
-
-def get_land_check_troubleshooting_message() -> str:
-    return (
-        " If you believe this is an error, you can use the old behavior with `@pytorchbot merge -g`"
-        + " (optionally with the `ciflow/trunk` to get land checks)"
-        + ' or use `@pytorchbot merge -f "some reason here"`.'
-        + f" For more information, see the [bot wiki]({BOT_COMMANDS_WIKI}). \n\n"
-        + CONTACT_US
-    )

.github/scripts/tryrebase.py

@@ -1,26 +1,22 @@
 #!/usr/bin/env python3
 
 import os
-import subprocess
-import sys
-import re
 from typing import Any
 from gitutils import get_git_remote_name, get_git_repo_dir, GitRepo
-from trymerge import gh_post_pr_comment as gh_post_comment, GitHubPR
+from trymerge import gh_post_comment, GitHubPR
 
 
 def parse_args() -> Any:
     from argparse import ArgumentParser
     parser = ArgumentParser("Rebase PR into branch")
     parser.add_argument("--dry-run", action="store_true")
-    parser.add_argument("--branch", type=str)
     parser.add_argument("pr_num", type=int)
     return parser.parse_args()
 
 
-def rebase_onto(pr: GitHubPR, repo: GitRepo, onto_branch: str, dry_run: bool = False) -> None:
+def rebase_onto(pr: GitHubPR, repo: GitRepo, dry_run: bool = False) -> None:
     branch = f"pull/{pr.pr_num}/head"
-    onto_branch = f"refs/remotes/origin/{onto_branch}"
+    onto_branch = pr.default_branch()
     remote_url = f"https://github.com/{pr.info['headRepository']['nameWithOwner']}.git"
     refspec = f"{branch}:{pr.head_ref()}"
 
@@ -40,91 +36,24 @@ def rebase_onto(pr: GitHubPR, repo: GitRepo, onto_branch: str, dry_run: bool = F
                         "git pull --rebase`)", dry_run=dry_run)
 
 
-def rebase_ghstack_onto(pr: GitHubPR, repo: GitRepo, onto_branch: str, dry_run: bool = False) -> None:
-    if subprocess.run([sys.executable, "-m", "ghstack", "--help"], capture_output=True).returncode != 0:
-        subprocess.run([sys.executable, "-m", "pip", "install", "ghstack"])
-    orig_ref = f"{re.sub(r'/head$', '/orig', pr.head_ref())}"
-    onto_branch = f"refs/remotes/origin/{onto_branch}"
-
-    repo.fetch(orig_ref, orig_ref)
-    repo._run_git("rebase", onto_branch, orig_ref)
-
-    # steal the identity of the committer of the commit on the orig branch
-    email = repo._run_git("log", orig_ref, "--pretty=format:%ae", "-1")
-    name = repo._run_git("log", orig_ref, "--pretty=format:%an", "-1")
-    repo._run_git("config", "--global", "user.name", name)
-    repo._run_git("config", "--global", "user.email", email)
-
-    os.environ["OAUTH_TOKEN"] = os.environ["GITHUB_TOKEN"]
-    with open('.ghstackrc', 'w+') as f:
-        f.write('[ghstack]\n' +
-                "github_url=github.com\n" +
-                "github_username=pytorchmergebot\n" +
-                "remote_name=origin")
-
-    if dry_run:
-        print("Don't know how to dry-run ghstack")
-    else:
-        ghstack_result = subprocess.run(["ghstack"], capture_output=True)
-        push_result = ghstack_result.stdout.decode("utf-8")
-        print(push_result)
-        if ghstack_result.returncode != 0:
-            raise Exception(f"\n```{push_result}```")
-        # The contents of a successful push result should look like:
-        # Summary of changes (ghstack 0.6.0)
-
-        #  - Updated https://github.com/clee2000/random-testing/pull/2
-        #  - Updated https://github.com/clee2000/random-testing/pull/1
-
-        # Facebook employees can import your changes by running
-        # (on a Facebook machine):
-
-        #     ghimport -s https://github.com/clee2000/random-testing/pull/2
-
-        # If you want to work on this diff stack on another machine:
-
-        #     ghstack checkout https://github.com/clee2000/random-testing/pull/2
-        org, project = repo.gh_owner_and_name()
-        for line in push_result.splitlines():
-            if "Updated" in line:
-                pr_num = int(line.split("/")[-1])
-                if pr_num != pr.pr_num:
-                    gh_post_comment(pr.org, pr.project, pr_num,
-                                    f"Rebased `{orig_ref}` onto `{onto_branch}` because #{pr.pr_num} was rebased, "
-                                    "please pull locally before adding more changes (for example, via `ghstack " +
-                                    f"checkout https://github.com/{org}/{project}/pull/{pr_num}`)", dry_run=dry_run)
-                else:
-                    gh_post_comment(pr.org, pr.project, pr_num,
-                                    f"Successfully rebased `{orig_ref}` onto `{onto_branch}`, please pull locally " +
-                                    "before adding more changes (for example, via `ghstack " +
-                                    f"checkout https://github.com/{org}/{project}/pull/{pr.pr_num}`)", dry_run=dry_run)
-
-        if f"Skipped https://github.com/{org}/{project}/pull/{pr.pr_num}" in push_result:
-            gh_post_comment(pr.org, pr.project, pr.pr_num,
-                            f"Tried to rebase and push PR #{pr.pr_num}, but it was already up to date", dry_run=dry_run)
-
-
 def main() -> None:
     args = parse_args()
     repo = GitRepo(get_git_repo_dir(), get_git_remote_name(), debug=True)
     org, project = repo.gh_owner_and_name()
 
     pr = GitHubPR(org, project, args.pr_num)
-    onto_branch = args.branch if args.branch else pr.default_branch()
-
-    msg = "@pytorchbot successfully started a rebase job."
-    msg += f" Check the current status [here]({os.getenv('GH_RUN_URL')})"
-    gh_post_comment(org, project, args.pr_num, msg, dry_run=args.dry_run)
 
     if pr.is_closed():
         gh_post_comment(org, project, args.pr_num, f"PR #{args.pr_num} is closed, won't rebase", dry_run=args.dry_run)
         return
 
+    if pr.is_ghstack_pr():
+        gh_post_comment(org, project, args.pr_num,
+                        f"PR #{args.pr_num} is a ghstack, which is currently not supported", dry_run=args.dry_run)
+        return
+
     try:
-        if pr.is_ghstack_pr():
-            rebase_ghstack_onto(pr, repo, onto_branch, dry_run=args.dry_run)
-            return
-        rebase_onto(pr, repo, onto_branch, dry_run=args.dry_run)
+        rebase_onto(pr, repo, dry_run=args.dry_run)
     except Exception as e:
         msg = f"Rebase failed due to {e}"
         run_url = os.getenv("GH_RUN_URL")

.github/scripts/update_commit_hashes.py

@@ -1,169 +0,0 @@
-import json
-import os
-import subprocess
-import requests
-from typing import Any, Dict
-from argparse import ArgumentParser
-
-MERGEBOT_TOKEN = os.environ["MERGEBOT_TOKEN"]
-PYTORCHBOT_TOKEN = os.environ["PYTORCHBOT_TOKEN"]
-OWNER, REPO = "pytorch", "pytorch"
-
-
-def git_api(
-    url: str, params: Dict[str, str], type: str = "get", token: str = MERGEBOT_TOKEN
-) -> Any:
-    headers = {
-        "Accept": "application/vnd.github.v3+json",
-        "Authorization": f"token {token}",
-    }
-    if type == "post":
-        return requests.post(
-            f"https://api.github.com{url}",
-            data=json.dumps(params),
-            headers=headers,
-        ).json()
-    elif type == "patch":
-        return requests.patch(
-            f"https://api.github.com{url}",
-            data=json.dumps(params),
-            headers=headers,
-        ).json()
-    else:
-        return requests.get(
-            f"https://api.github.com{url}",
-            params=params,
-            headers=headers,
-        ).json()
-
-
-def parse_args() -> Any:
-    parser = ArgumentParser("Rebase PR into branch")
-    parser.add_argument("--repo-name", type=str)
-    parser.add_argument("--branch", type=str)
-    return parser.parse_args()
-
-
-def make_pr(repo_name: str, branch_name: str) -> Any:
-    params = {
-        "title": f"[{repo_name} hash update] update the pinned {repo_name} hash",
-        "head": branch_name,
-        "base": "master",
-        "body": "This PR is auto-generated nightly by [this action](https://github.com/pytorch/pytorch/blob/master/"
-        + f".github/workflows/_update-commit-hash.yml).\nUpdate the pinned {repo_name} hash.",
-    }
-    response = git_api(f"/repos/{OWNER}/{REPO}/pulls", params, type="post")
-    print(f"made pr {response['html_url']}")
-    return response["number"]
-
-
-def approve_pr(pr_number: str) -> None:
-    params = {"event": "APPROVE"}
-    # use pytorchbot to approve the pr
-    git_api(
-        f"/repos/{OWNER}/{REPO}/pulls/{pr_number}/reviews",
-        params,
-        type="post",
-        token=PYTORCHBOT_TOKEN,
-    )
-
-
-def make_comment(pr_number: str, msg: str) -> None:
-    params = {"body": msg}
-    # comment with pytorchbot because pytorchmergebot gets ignored
-    git_api(
-        f"/repos/{OWNER}/{REPO}/issues/{pr_number}/comments",
-        params,
-        type="post",
-        token=PYTORCHBOT_TOKEN,
-    )
-
-
-def close_pr(pr_number: str) -> None:
-    params = {"state": "closed"}
-    git_api(
-        f"/repos/{OWNER}/{REPO}/pulls/{pr_number}",
-        params,
-        type="patch",
-    )
-
-
-def is_newer_hash(new_hash: str, old_hash: str, repo_name: str) -> bool:
-    def _get_date(hash: str) -> int:
-        # this git command prints the unix timestamp of the hash
-        return int(
-            subprocess.run(
-                f"git show --no-patch --no-notes --pretty=%ct {hash}".split(),
-                capture_output=True,
-                cwd=f"{repo_name}",
-            )
-            .stdout.decode("utf-8")
-            .strip()
-        )
-
-    return _get_date(new_hash) > _get_date(old_hash)
-
-
-def main() -> None:
-    args = parse_args()
-
-    branch_name = os.environ["NEW_BRANCH_NAME"]
-    pr_num = None
-
-    # query to see if a pr already exists
-    params = {
-        "q": f"is:pr is:open in:title author:pytorchmergebot repo:{OWNER}/{REPO} {args.repo_name} hash update"
-    }
-    response = git_api("/search/issues", params)
-    if response["total_count"] != 0:
-        # pr does exist
-        pr_num = response["items"][0]["number"]
-        link = response["items"][0]["html_url"]
-        response = git_api(f"/repos/{OWNER}/{REPO}/pulls/{pr_num}", {})
-        branch_name = response["head"]["ref"]
-        print(
-            f"pr does exist, number is {pr_num}, branch name is {branch_name}, link is {link}"
-        )
-
-    hash = (
-        subprocess.run(
-            f"git rev-parse {args.branch}".split(),
-            capture_output=True,
-            cwd=f"{args.repo_name}",
-        )
-        .stdout.decode("utf-8")
-        .strip()
-    )
-    with open(f".github/ci_commit_pins/{args.repo_name}.txt", "r+") as f:
-        old_hash = f.read().strip()
-        subprocess.run(f"git checkout {old_hash}".split(), cwd=args.repo_name)
-        f.seek(0)
-        f.truncate()
-        f.write(f"{hash}\n")
-    if is_newer_hash(hash, old_hash, args.repo_name):
-        # if there was an update, push to branch
-        subprocess.run(f"git checkout -b {branch_name}".split())
-        subprocess.run(f"git add .github/ci_commit_pins/{args.repo_name}.txt".split())
-        subprocess.run(
-            "git commit -m".split() + [f"update {args.repo_name} commit hash"]
-        )
-        subprocess.run(f"git push --set-upstream origin {branch_name} -f".split())
-        print(f"changes pushed to branch {branch_name}")
-        if pr_num is None:
-            # no existing pr, so make a new one and approve it
-            pr_num = make_pr(args.repo_name, branch_name)
-            approve_pr(pr_num)
-        # comment to merge if all checks are green
-        make_comment(pr_num, "@pytorchbot merge -g")
-    else:
-        print(
-            f"tried to update from old hash: {old_hash} to new hash: {hash} but the old hash seems to be newer, not creating pr"
-        )
-        if pr_num is not None:
-            make_comment(pr_num, "closing pr as the current hash seems up to date")
-            close_pr(pr_num)
-            print(f"closing PR {pr_num}")
-
-
-if __name__ == "__main__":
-    main()

.github/scripts/wait_for_ssh_to_drain.sh

@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+
+set -eou pipefail
+
+echo "Holding runner for 2 hours until all ssh sessions have logged out"
+for _ in $(seq 1440); do
+    # Break if no ssh session exists anymore
+    if [ "$(who)" = "" ]; then
+      break
+    fi
+    echo "."
+    sleep 5
+done

.github/templates/common.yml.j2

@@ -1,7 +1,4 @@
-{%- set upload_artifact_s3_action = "seemethere/upload-artifact-s3@v5" -%}
-{%- set download_artifact_s3_action = "seemethere/download-artifact-s3@v4" -%}
-{%- set upload_artifact_action = "actions/upload-artifact@v3" -%}
-{%- set download_artifact_action = "actions/download-artifact@v3" -%}
+{%- set upload_artifact_s3_action = "seemethere/upload-artifact-s3@v4" -%}
 
 {# squid_proxy is an private ELB that only available for GHA custom runners #}
 {%- set squid_proxy    = "http://internal-tf-lb-20210727220640487900000002-835786077.us-east-1.elb.amazonaws.com:3128" -%}
@@ -11,11 +8,11 @@
 
 # NOTE: If testing pytorch/builder changes you can change this variable to change what pytorch/builder reference
 #       the binary builds will check out
-{%- set builder_branch = "release/1.13" -%}
+{%- set builder_branch = "release/1.12" -%}
 
 {%- macro concurrency(build_environment) -%}
 concurrency:
-  group: !{{ build_environment }}-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
+  group: !{{ build_environment }}-${{ github.event.pull_request.number || github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
   cancel-in-progress: true
 {%- endmacro -%}
 
@@ -93,6 +90,7 @@ on:
           AWS_DEFAULT_REGION: us-east-1
           GIT_DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
           BRANCH: ${{ steps.parse-ref.outputs.branch }}
+          JOB_BASE_NAME: !{{ build_environment }}-test
           PR_NUMBER: ${{ github.event.pull_request.number }}
           SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
           TAG: ${{ steps.parse-ref.outputs.tag }}
@@ -195,8 +193,30 @@ on:
           killall runsvc.sh
       - name: Preserve github env variables for use in docker
         run: |
-          env | grep '^GITHUB' >> "/tmp/github_env_${GITHUB_RUN_ID}"
-          env | grep '^CI' >> "/tmp/github_env_${GITHUB_RUN_ID}"
+          env | grep '^GITHUB' > "/tmp/github_env_${GITHUB_RUN_ID}"
+{%- endmacro -%}
+
+{%- macro teardown_ec2_linux(pytorch_directory="") -%}
+      - name: Hold runner for 2 hours or until ssh sessions have drained
+{%- if pytorch_directory %}
+        working-directory: !{{ pytorch_directory }}
+{%- endif %}
+        # Always hold for active ssh sessions
+        if: always()
+        run: .github/scripts/wait_for_ssh_to_drain.sh
+      - name: Chown workspace
+        if: always()
+        run: |
+          # Ensure the working directory gets chowned back to the current user
+          docker run --rm -v "$(pwd)":/v -w /v "${ALPINE_IMAGE}" chown -R "$(id -u):$(id -g)" .
+      - name: Kill containers, clean up images
+        if: always()
+        run: |
+          # ignore expansion of "docker ps -q" since it could be empty
+          # shellcheck disable=SC2046
+          docker stop $(docker ps -q) || true
+          # Prune all of the docker images
+          docker system prune -af
 {%- endmacro -%}
 
 {%- macro teardown_rocm_linux() -%}