Merge branch 'main' into x64_workflow_test

Author: alinpahontu2912

.git-blame-ignore-revs

@@ -11,3 +11,5 @@ d367a01a18a3ae6bee13d8be3b63fd6a581ea46f
 a335d916db0694770e8152f41e19195de3134523
 # Renaming: `BoundingBox` -> `BoundingBoxes` (#7778)
 332bff937c6711666191880fab57fa2f23ae772e
+# Upgrade type hint and others to Python 3.9 (#8814)
+a095de183d3811d79ed0db2715e7a1c3162fa19d

.github/workflows/build-wheels-aarch64-linux.yml

@@ -25,7 +25,7 @@ jobs:
       os: linux-aarch64
       test-infra-repository: pytorch/test-infra
       test-infra-ref: main
-      with-cuda: disable
+      with-cuda: enable
   build:
     needs: generate-matrix
     strategy:

docs/source/io.rst

@@ -41,7 +41,7 @@ powerful, e.g. if you want to encode/decode JPEGs on CUDA.
 
     decode_image
     decode_jpeg
-    encode_png
+    decode_png
     decode_webp
     decode_avif
     decode_heic

docs/source/training_references.rst

@@ -19,9 +19,9 @@ guarantees.
 
 In general, these scripts rely on the latest (not yet released) pytorch version
 or the latest torchvision version. This means that to use them, **you might need
-to install the latest pytorch and torchvision versions**, with e.g.::
+to install the latest pytorch and torchvision versions** following the `official
+instructions <https://pytorch.org/get-started/locally/>`_.
 
-    conda install pytorch torchvision -c pytorch-nightly
 
 If you need to rely on an older stable version of pytorch or torchvision, e.g.
 torchvision 0.10, then it's safer to use the scripts from that corresponding

mypy.ini

@@ -29,6 +29,10 @@ ignore_errors = True
 
 ignore_errors = True
 
+[mypy-torchvision.prototype.models.*]
+
+ignore_errors = True
+
 [mypy-torchvision.io.image.*]
 
 ignore_errors = True
@@ -41,54 +45,10 @@ ignore_errors = True
 
 ignore_errors = True
 
-[mypy-torchvision.models.densenet.*]
+[mypy-torchvision.models.*]
 
 ignore_errors=True
 
-[mypy-torchvision.models.maxvit.*]
-
-ignore_errors=True
-
-[mypy-torchvision.models.detection.anchor_utils]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.transform]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.roi_heads]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.faster_rcnn]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.mask_rcnn]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.keypoint_rcnn]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.retinanet]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.ssd]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.ssdlite]
-
-ignore_errors = True
-
-[mypy-torchvision.models.detection.fcos]
-
-ignore_errors = True
-
 [mypy-torchvision.ops.*]
 
 ignore_errors = True

packaging/pre_build_script_arm64.sh

@@ -27,6 +27,7 @@ git clone https://github.com/microsoft/vcpkg.git
 cd vcpkg || exit
 ./bootstrap-vcpkg.sh
 
+
 # # Set vcpkg to only build release packages
 echo "set(VCPKG_BUILD_TYPE release)" >> "$TRIPLET_FILE"
 
@@ -41,6 +42,7 @@ cp "$DEPENDENCIES_DIR/x64-windows/bin/libpng16.dll" "$DEPENDENCIES_DIR/x64-windo
 cp "$DEPENDENCIES_DIR/x64-windows/bin/libpng16.pdb" "$DEPENDENCIES_DIR/x64-windows/bin/libpng.pdb"
 mkdir -p "$DEPENDENCIES_DIR/Library/"
 cp -r "$DEPENDENCIES_DIR/x64-windows/"* "$DEPENDENCIES_DIR/Library/"
+
 cp -r "$DEPENDENCIES_DIR/Library/tools/libpng/"* "$DEPENDENCIES_DIR/Library/bin/"
 cp -r "$DEPENDENCIES_DIR/Library/bin/"* "$SRC_PATH/torchvision"
 
@@ -58,22 +60,15 @@ pip install numpy==2.2.3
 
 if [ "$CHANNEL" = "release" ]; then
   echo "Installing latest stable version of PyTorch."
+
   pip3 install torch
 elif [ "$CHANNEL" = "test" ]; then
   echo "Installing PyTorch version $PYTORCH_VERSION."
   pip3 install torch=="$PYTORCH_VERSION"
+
 else
   echo "CHANNEL is not set, installing PyTorch from nightly."
   pip3 install --pre torch --index-url https://download.pytorch.org/whl/nightly/cpu
 fi
 
-# # Create wheel under dist folder
-# python setup.py bdist_wheel
-
-# # Check if build was successful
-# if [[ $? -ne 0 ]]; then
-#     echo "Failed on build_vision. (exit code = $?)"
-#     exit 1
-# fi
-
 echo "Dependencies install finished successfully."

setup.py

@@ -24,7 +24,7 @@
 NVCC_FLAGS = os.getenv("NVCC_FLAGS", None)
 # Note: the GPU video decoding stuff used to be called "video codec", which
 # isn't an accurate or descriptive name considering there are at least 2 other
-# video deocding backends in torchvision. I'm renaming this to "gpu video
+# video decoding backends in torchvision. I'm renaming this to "gpu video
 # decoder" where possible, keeping user facing names (like the env var below) to
 # the old scheme for BC.
 USE_GPU_VIDEO_DECODER = os.getenv("TORCHVISION_USE_VIDEO_CODEC", "1") == "1"
@@ -211,7 +211,7 @@ def find_libpng():
             subprocess.run([libpng_config, "--version"], stdout=subprocess.PIPE).stdout.strip().decode("utf-8")
         )
         if png_version < min_version:
-            warnings.warn("libpng version {png_version} is less than minimum required version {min_version}")
+            warnings.warn(f"libpng version {png_version} is less than minimum required version {min_version}")
             return False, None, None, None
 
         include_dir = (
@@ -271,6 +271,14 @@ def find_library(header):
                 return True, None, None
             print(f"{searching_for}. Didn't find in {prefix}")
 
+    if sys.platform == "darwin":
+        HOMEBREW_PATH = Path("/opt/homebrew")
+        include_dir = HOMEBREW_PATH / "include"
+        library_dir = HOMEBREW_PATH / "lib"
+        if (include_dir / header).exists():
+            print(f"{searching_for}. Found in {include_dir}.")
+            return True, str(include_dir), str(library_dir)
+
     return False, None, None
 
 
@@ -440,7 +448,7 @@ def find_ffmpeg_libraries():
 
         extensions.append(
             CppExtension(
-                # This is an aweful name. It should be "cpu_video_decoder". Keeping for BC.
+                # This is an awful name. It should be "cpu_video_decoder". Keeping for BC.
                 "torchvision.video_reader",
                 combined_src,
                 include_dirs=[

test/assets/fakedata/draw_boxes_different_label_fill_colors.png

@@ -1,7 +1,7 @@
 import os
 from collections import defaultdict
 from numbers import Number
-from typing import Any, List
+from typing import Any
 
 import torch
 from torch.utils._python_dispatch import TorchDispatchMode
@@ -30,7 +30,7 @@ def prod(x):
     return res
 
 
-def matmul_flop(inputs: List[Any], outputs: List[Any]) -> Number:
+def matmul_flop(inputs: list[Any], outputs: list[Any]) -> Number:
     """
     Count flops for matmul.
     """
@@ -43,7 +43,7 @@ def matmul_flop(inputs: List[Any], outputs: List[Any]) -> Number:
     return flop
 
 
-def addmm_flop(inputs: List[Any], outputs: List[Any]) -> Number:
+def addmm_flop(inputs: list[Any], outputs: list[Any]) -> Number:
     """
     Count flops for fully connected layers.
     """
@@ -60,7 +60,7 @@ def addmm_flop(inputs: List[Any], outputs: List[Any]) -> Number:
     return flops
 
 
-def bmm_flop(inputs: List[Any], outputs: List[Any]) -> Number:
+def bmm_flop(inputs: list[Any], outputs: list[Any]) -> Number:
     """
     Count flops for the bmm operation.
     """
@@ -75,9 +75,9 @@ def bmm_flop(inputs: List[Any], outputs: List[Any]) -> Number:
 
 
 def conv_flop_count(
-    x_shape: List[int],
-    w_shape: List[int],
-    out_shape: List[int],
+    x_shape: list[int],
+    w_shape: list[int],
+    out_shape: list[int],
     transposed: bool = False,
 ) -> Number:
     """
@@ -99,7 +99,7 @@ def conv_flop_count(
     return flop
 
 
-def conv_flop(inputs: List[Any], outputs: List[Any]):
+def conv_flop(inputs: list[Any], outputs: list[Any]):
     """
     Count flops for convolution.
     """
@@ -110,7 +110,7 @@ def conv_flop(inputs: List[Any], outputs: List[Any]):
     return conv_flop_count(x_shape, w_shape, out_shape, transposed=transposed)
 
 
-def quant_conv_flop(inputs: List[Any], outputs: List[Any]):
+def quant_conv_flop(inputs: list[Any], outputs: list[Any]):
     """
     Count flops for quantized convolution.
     """
@@ -124,8 +124,8 @@ def transpose_shape(shape):
     return [shape[1], shape[0]] + list(shape[2:])
 
 
-def conv_backward_flop(inputs: List[Any], outputs: List[Any]):
-    grad_out_shape, x_shape, w_shape = [get_shape(i) for i in inputs[:3]]
+def conv_backward_flop(inputs: list[Any], outputs: list[Any]):
+    grad_out_shape, x_shape, w_shape = (get_shape(i) for i in inputs[:3])
     output_mask = inputs[-1]
     fwd_transposed = inputs[7]
     flop_count = 0
@@ -140,7 +140,7 @@ def conv_backward_flop(inputs: List[Any], outputs: List[Any]):
     return flop_count
 
 
-def scaled_dot_product_flash_attention_flop(inputs: List[Any], outputs: List[Any]):
+def scaled_dot_product_flash_attention_flop(inputs: list[Any], outputs: list[Any]):
     # FIXME: this needs to count the flops of this kernel
     # https://github.com/pytorch/pytorch/blob/207b06d099def9d9476176a1842e88636c1f714f/aten/src/ATen/native/cpu/FlashAttentionKernel.cpp#L52-L267
     return 0