Publish Linux CPython 3.10 MMCV CPU wheel

Add unmodified MMCV 2.2.0 CPU wheel for Linux x86-64/glibc >=2.17 and Torch 2.4.0+cpu, original build reports, independent ELF/hash/license audit, host operator/model feature checks, and additive notices. Existing wheels and notices archives are unchanged. Clean-host full MAAP3D and checkpoint inference remain pending.

.gitattributes

@@ -65,3 +65,4 @@ sam3-0.1.0+maap3d.847e1a3b.2-py3-none-any.whl filter=lfs diff=lfs merge=lfs -tex
 mmcv-2.2.0-cp310-cp310-win_amd64.whl filter=lfs diff=lfs merge=lfs -text
 mmcv-2.2.0-1cu128sm120-cp310-cp310-win_amd64.whl filter=lfs diff=lfs merge=lfs -text
 mmcv-2.2.0-1cpupt241-cp310-cp310-win_amd64.whl filter=lfs diff=lfs merge=lfs -text
+mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl filter=lfs diff=lfs merge=lfs -text

PROVENANCE.json

@@ -245,6 +245,91 @@
         "builds/windows-cpu/DLL_DEPENDENCIES.txt",
         "builds/windows-cpu/PUBLICATION_CHECKS.json"
       ]
+    },
+    {
+      "key": "mmcv-linux-cpu-cp310",
+      "distribution": "mmcv",
+      "version": "2.2.0",
+      "build_tag": "1cpupt240",
+      "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+      "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+      "upstream_repository": "https://github.com/open-mmlab/mmcv",
+      "upstream_version": "2.2.0",
+      "upstream_source_url": "https://files.pythonhosted.org/packages/e9/a2/57a733e7e84985a8a0e3101dfb8170fc9db92435c16afad253069ae3f9df/mmcv-2.2.0.tar.gz",
+      "upstream_source_sha256": "ac479247e808d8802f89eadf04d4118de86bdfe81361ec5aed0cc1bf731c67c9",
+      "profile": "Linux/WSL2 x86-64 CPython 3.10 / Torch 2.4.0+cpu / glibc >=2.17 only",
+      "changes": [
+        "Compile unmodified MMCV 2.2.0 with CPU operators only (MMCV_WITH_OPS=1, FORCE_CUDA=0).",
+        "Distinct wheel build tag 1cpupt240; auditwheel manylinux2014 repair excludes matching external Torch libraries."
+      ],
+      "builder": "builds/linux-cpu/build_mmcv.py",
+      "compiler": "gcc (GCC) 10.2.1 20210130 (Red Hat 10.2.1-11)",
+      "build_image": "quay.io/pypa/manylinux2014_x86_64@sha256:55df5597db86b0a03d5f21a03b060271efffc78ef97123c60ee9c7ce0f10ec31",
+      "license": "Apache-2.0 generally; see operation-specific upstream LICENSES.md and existing NVIDIA notices",
+      "license_files": [
+        "licenses/mmcv/LICENSE",
+        "licenses/mmcv/LICENSES.md",
+        "licenses/mmcv/NVIDIA-StyleGAN2-LICENSE.txt",
+        "licenses/mmcv/NVIDIA-StyleGAN3-LICENSE.txt"
+      ],
+      "validation": {
+        "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+        "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+        "size_bytes": 30511921,
+        "zip_crc": "passed",
+        "duplicate_archive_members": false,
+        "record_hashes_and_sizes_verified": 537,
+        "elf": "ELF64 little-endian x86-64",
+        "extension": "mmcv/_ext.cpython-310-x86_64-linux-gnu.so",
+        "build_tag": "1cpupt240",
+        "wheel_tags": [
+          "cp310-cp310-manylinux2014_x86_64",
+          "cp310-cp310-manylinux_2_17_x86_64"
+        ],
+        "needed_libraries": [
+          "libc10.so",
+          "libtorch.so",
+          "libtorch_cpu.so",
+          "libtorch_python.so",
+          "libstdc++.so.6",
+          "libm.so.6",
+          "libgcc_s.so.1",
+          "libpthread.so.0",
+          "libc.so.6"
+        ],
+        "cuda_library_dependencies": false,
+        "rpath_or_runpath": false,
+        "highest_extension_glibc_symbol": "2.14",
+        "licenses": [
+          "mmcv-2.2.0.dist-info/licenses/",
+          "mmcv-2.2.0.dist-info/licenses/LICENSE",
+          "mmcv-2.2.0.dist-info/licenses/LICENSES.md"
+        ],
+        "license_texts_match_existing_publication": true,
+        "original_build_validation": {
+          "cpu_nms": "passed",
+          "cpu_roi_align_forward_backward": "passed",
+          "mmpose_hrnet_features": "passed",
+          "mmdet_maskrcnn_features": "passed",
+          "pip_check": "No broken requirements found."
+        },
+        "publication_host_smoke": "CPU NMS, RoIAlign forward/backward, checkpoint-free MMPose HRNet and MMDetection Mask R-CNN features passed using exact candidate installed in isolated temporary target over existing CPU runtime",
+        "publication_host_pip_check": "passed (temporary candidate with existing CPU dependencies)",
+        "not_run": [
+          "Complete clean-host compiler-free MAAP3D installation",
+          "Checkpoint-backed inference"
+        ],
+        "abi_note": "Requires matching external Torch 2.4.0+cpu libraries; not a standalone manylinux distribution"
+      },
+      "reports": [
+        "builds/linux-cpu/BUILD_REPORT.json",
+        "builds/linux-cpu/INTEGRITY.json",
+        "builds/linux-cpu/AUDITWHEEL.txt",
+        "builds/linux-cpu/SMOKE_TEST.txt",
+        "builds/linux-cpu/PUBLICATION_CHECKS.json",
+        "builds/linux-cpu/PUBLICATION_SMOKE.txt",
+        "builds/linux-cpu/PUBLICATION_PIP_CHECK.txt"
+      ]
     }
   ],
   "builder": "tools/build_patched_wheels.py",

README.md

@@ -30,13 +30,14 @@ redistributing or using an artifact.
 | `mmcv` | `2.2.0`, Windows `cp310` | Torch 2.4.1/cu121; SM75/80/86/89 | Apache-2.0 generally; see MMCV special-operation NVIDIA notices |
 | `mmcv` | `2.2.0`, build `1cu128sm120`, Windows `cp310` | Experimental Torch 2.7.1/cu128; SM120 only | Apache-2.0 generally; see MMCV special-operation NVIDIA notices |
 | `mmcv` | `2.2.0`, build `1cpupt241`, Windows `cp310` | Experimental Torch 2.4.1+cpu; AMD64 | Apache-2.0 generally; see MMCV operation-specific notices |
+| `mmcv` | `2.2.0`, build `1cpupt240`, Linux `cp310` | Torch 2.4.0+cpu; x86-64 / glibc >=2.17 | Apache-2.0 generally; see MMCV operation-specific notices |
 | `mmdet` | `3.3.0+maap3d.cfd5d3a9.1` | CUDA 12.1 and CUDA 12.8 full/dev | Apache-2.0 |
 | `sam3` | `0.1.0+maap3d.847e1a3b.1` | Legacy CUDA 12.8 artifact with PyTorch >=2.7 | Custom SAM License dated November 19, 2025 |
 | `sam3` | `0.1.0+maap3d.847e1a3b.2` | CUDA 12.1 and CUDA 12.8 full/dev with PyTorch >=2.4 | Custom SAM License dated November 19, 2025 |
 | `chumpy` | `0.71+maap3d.580566ea.1` | MAAP3D OpenMMLab stacks, including Python 3.12 GB10 | MIT |
 | `mmpose` | `1.3.2+maap3d.gb10.5408bc76.1` | GB10 `mmcv-lite` only | Apache-2.0 generally; EDPose files are identified as IDEA License 1.0 |
 
-This repository contains pure-Python wheels and compiled Windows MMCV wheels.
+This repository contains pure-Python wheels and compiled Windows and Linux CPU MMCV wheels.
 Each artifact is specific to its stated Python, PyTorch, CUDA, and platform
 profile. Do not install every artifact together as a generic environment.
 
@@ -104,6 +105,38 @@ inference, and a clean Windows machine without Build Tools have NOT RUN.**
 The installer retains its CPU operator/model smoke and final `pip check`.
 This remains an experimental component release pending those checks.
 
+## Linux CPU MMCV wheel
+
+`mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl`
+is built from unmodified MMCV 2.2.0 for CPython 3.10, Linux/WSL2 x86-64,
+glibc >=2.17 and exactly Torch 2.4.0+cpu (TorchVision 0.19.0+cpu /
+TorchAudio 2.4.0+cpu). SHA-256:
+`d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970`.
+
+This wheel contains CPU operators and no CUDA dependencies. Matching Torch
+libraries remain external: the manylinux tags do not make it a standalone
+manylinux distribution or encode the Torch ABI. Python 3.11, ARM64,
+32-bit Python, musl/Alpine, and CUDA Torch are outside this artifact's target.
+Import Torch before directly loading MMCV operators.
+
+The manylinux2014/glibc 2.17 build and fresh verification environment passed
+CPU NMS, RoIAlign forward/backward, checkpoint-free MMPose HRNet and
+MMDetection Mask R-CNN feature checks, and `pip check`. Independent publication
+review checked SHA-256, ZIP CRC, 537 RECORD hashes/sizes, ELF64 x86-64 identity,
+licenses and all shared-library dependencies. Those runtime smokes and
+`pip check` also passed on the Linux publication host using the exact wheel
+installed into a temporary target over the existing CPU runtime. Build scripts,
+original logs and publication checks are under
+[`builds/linux-cpu/`](builds/linux-cpu/). Original candidate-only report wording
+records build-time status before this publication.
+
+The updated MAAP3D `python install_package.py --cpu` selects the wheel at an
+immutable revision, verifies SHA-256, and force-reinstalls with `--no-deps`.
+It checks Python, architecture and glibc compatibility before changing packages.
+No MMCV source compilation, CUDA toolkit or NVIDIA GPU is required.
+**Complete installation on a clean host without build tools and real
+checkpoint-backed inference remain NOT RUN.**
+
 ## What was changed
 
 - **MMDetection:** widens the package's runtime MMCV version guard so the
@@ -174,15 +207,18 @@ After downloading the complete artifact set, verify it with:
 sha256sum -c SHA256SUMS
 ```
 
-`SHA256SUMS` covers the eight wheels, the original
+`SHA256SUMS` covers the nine wheels, the original
 `THIRD_PARTY_WHEEL_NOTICES.zip`, and the additive
-`THIRD_PARTY_WHEEL_NOTICES-windows-cpu-20261025.zip`. It does not cover repository documentation,
+`THIRD_PARTY_WHEEL_NOTICES-windows-cpu-20261025.zip` and
+`THIRD_PARTY_WHEEL_NOTICES-linux-cpu-20261025.zip`. It does not cover repository documentation,
 the standalone license copies, or `PROVENANCE.json`.
 
 The original notices archive retains its earlier release snapshot. The additive
 [Windows CPU notices archive](THIRD_PARTY_WHEEL_NOTICES-windows-cpu-20261025.zip)
 contains the updated release notice/provenance and applicable MMCV license
-copies, without replacing the existing archive. Each wheel also carries its applicable package license
+copies, without replacing the existing archive. The additive
+[Linux CPU notices archive](THIRD_PARTY_WHEEL_NOTICES-linux-cpu-20261025.zip)
+contains the subsequent release notice/provenance and applicable MMCV licenses. Each wheel also carries its applicable package license
 metadata. Model checkpoints, PyTorch/CUDA runtimes, `xtcocotools`, and
 other transitive dependencies are not included. Model checkpoints can have
 terms separate from their source code and must be obtained from an authorized

SHA256SUMS

@@ -8,3 +8,5 @@ c67bca0e3d24611cf4241150f5ce9e03ca8b8697f38bb2ba497cabbf996cfcdd  mmcv-2.2.0-1cu
 14b7d2bed1ae109608b1e50b722c6cbe13ea6bdf01e1cbe1aa57e1a56fa8d4ae  THIRD_PARTY_WHEEL_NOTICES.zip
 35db4db7d467f45e71e4d50c19ffaf6e35c713053109ff413d199ac503a45389  mmcv-2.2.0-1cpupt241-cp310-cp310-win_amd64.whl
 99d8adfd9a317a41b086e3bcf2fc82bed0d6ab0a977fa3ba6853d84aaa3b04a9  THIRD_PARTY_WHEEL_NOTICES-windows-cpu-20261025.zip
+d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970  mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
+2b6af5946b18506f5ab5e9d6b69324e6f2aff810b0e011a681dbb489d08c16a9  THIRD_PARTY_WHEEL_NOTICES-linux-cpu-20261025.zip

THIRD_PARTY_WHEEL_NOTICES-linux-cpu-20261025.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b6af5946b18506f5ab5e9d6b69324e6f2aff810b0e011a681dbb489d08c16a9
+size 20031

builds/linux-cpu/AUDITWHEEL.txt

@@ -0,0 +1,16 @@
+
+mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
+is consistent with the following platform tag: "linux_x86_64".
+
+The wheel references external versioned symbols in these
+system-provided shared libraries: libpthread.so.0 with versions
+{'GLIBC_2.2.5'}, libgcc_s.so.1 with versions {'GCC_3.0'}, libm.so.6
+with versions {'GLIBC_2.2.5'}, libc.so.6 with versions {'GLIBC_2.2.5',
+'GLIBC_2.14'}, libstdc++.so.6 with versions {'CXXABI_1.3.3',
+'CXXABI_1.3.2', 'CXXABI_1.3.5', 'GLIBCXX_3.4.9', 'GLIBCXX_3.4.18',
+'GLIBCXX_3.4', 'CXXABI_1.3'}
+
+This constrains the platform tag to "manylinux_2_17_x86_64". In order
+to achieve a more compatible tag, you would need to recompile a new
+wheel from source on a system with earlier versions of these
+libraries, such as a recent manylinux image.

builds/linux-cpu/BUILD_REPORT.json

@@ -0,0 +1,51 @@
+{
+  "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+  "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+  "size_bytes": 30511921,
+  "python": "3.10.22",
+  "architecture": "x86_64",
+  "glibc": [
+    "glibc",
+    "2.17"
+  ],
+  "compiler": "gcc (GCC) 10.2.1 20210130 (Red Hat 10.2.1-11)",
+  "torch": "2.4.0+cpu",
+  "numpy": "1.26.4",
+  "mmcv": "2.2.0",
+  "source": {
+    "filename": "mmcv-2.2.0.tar.gz",
+    "sha256": "ac479247e808d8802f89eadf04d4118de86bdfe81361ec5aed0cc1bf731c67c9",
+    "url": "https://files.pythonhosted.org/packages/e9/a2/57a733e7e84985a8a0e3101dfb8170fc9db92435c16afad253069ae3f9df/mmcv-2.2.0.tar.gz"
+  },
+  "extension": "mmcv/_ext.cpython-310-x86_64-linux-gnu.so",
+  "external_torch_libraries": [
+    "libc10.so",
+    "libtorch.so",
+    "libtorch_cpu.so",
+    "libtorch_python.so"
+  ],
+  "build_image": "quay.io/pypa/manylinux2014_x86_64@sha256:55df5597db86b0a03d5f21a03b060271efffc78ef97123c60ee9c7ce0f10ec31",
+  "validation": {
+    "cpu_nms": "passed",
+    "cpu_roi_align_forward_backward": "passed",
+    "mmpose_hrnet_features": "passed",
+    "mmdet_maskrcnn_features": "passed",
+    "pip_check": "No broken requirements found."
+  },
+  "limits": [
+    "CPython 3.10 and Linux x86-64 only; no CUDA operators",
+    "Requires the matching PyTorch CPU libraries; not a standalone manylinux wheel",
+    "No model checkpoints or complete MAAP3D installation tested",
+    "Local candidate; not published or wired into installers"
+  ],
+  "build_backend": "distutils fallback; future recipe adds Ninja to PATH",
+  "integrity_report": "INTEGRITY.json: 537 verified RECORD hashes; no CUDA shared-library dependencies",
+  "build_requirements": {
+    "pip": "25.0.1",
+    "setuptools": "80.9.0",
+    "wheel": "0.44.0",
+    "ninja": "1.11.1.4",
+    "psutil": "6.1.1",
+    "auditwheel": "6.4.2"
+  }
+}

builds/linux-cpu/INTEGRITY.json

@@ -0,0 +1,21 @@
+{
+  "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+  "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+  "record_hashes_verified": 537,
+  "licenses": [
+    "mmcv-2.2.0.dist-info/licenses/LICENSE",
+    "mmcv-2.2.0.dist-info/licenses/LICENSES.md"
+  ],
+  "needed_libraries": [
+    "0x0000000000000001 (NEEDED)             Shared library: [libc10.so]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libtorch.so]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libtorch_cpu.so]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libtorch_python.so]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libstdc++.so.6]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libm.so.6]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libgcc_s.so.1]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libpthread.so.0]",
+    "0x0000000000000001 (NEEDED)             Shared library: [libc.so.6]"
+  ],
+  "cuda_shared_library_dependencies": false
+}

builds/linux-cpu/PIP_CHECK.txt

@@ -0,0 +1 @@
+No broken requirements found.

builds/linux-cpu/PUBLICATION_CHECKS.json

@@ -0,0 +1,49 @@
+{
+  "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+  "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+  "size_bytes": 30511921,
+  "zip_crc": "passed",
+  "duplicate_archive_members": false,
+  "record_hashes_and_sizes_verified": 537,
+  "elf": "ELF64 little-endian x86-64",
+  "extension": "mmcv/_ext.cpython-310-x86_64-linux-gnu.so",
+  "build_tag": "1cpupt240",
+  "wheel_tags": [
+    "cp310-cp310-manylinux2014_x86_64",
+    "cp310-cp310-manylinux_2_17_x86_64"
+  ],
+  "needed_libraries": [
+    "libc10.so",
+    "libtorch.so",
+    "libtorch_cpu.so",
+    "libtorch_python.so",
+    "libstdc++.so.6",
+    "libm.so.6",
+    "libgcc_s.so.1",
+    "libpthread.so.0",
+    "libc.so.6"
+  ],
+  "cuda_library_dependencies": false,
+  "rpath_or_runpath": false,
+  "highest_extension_glibc_symbol": "2.14",
+  "licenses": [
+    "mmcv-2.2.0.dist-info/licenses/",
+    "mmcv-2.2.0.dist-info/licenses/LICENSE",
+    "mmcv-2.2.0.dist-info/licenses/LICENSES.md"
+  ],
+  "license_texts_match_existing_publication": true,
+  "original_build_validation": {
+    "cpu_nms": "passed",
+    "cpu_roi_align_forward_backward": "passed",
+    "mmpose_hrnet_features": "passed",
+    "mmdet_maskrcnn_features": "passed",
+    "pip_check": "No broken requirements found."
+  },
+  "publication_host_smoke": "CPU NMS, RoIAlign forward/backward, checkpoint-free MMPose HRNet and MMDetection Mask R-CNN features passed using exact candidate installed in isolated temporary target over existing CPU runtime",
+  "publication_host_pip_check": "passed (temporary candidate with existing CPU dependencies)",
+  "not_run": [
+    "Complete clean-host compiler-free MAAP3D installation",
+    "Checkpoint-backed inference"
+  ],
+  "abi_note": "Requires matching external Torch 2.4.0+cpu libraries; not a standalone manylinux distribution"
+}

builds/linux-cpu/PUBLICATION_INSTALL.txt

@@ -0,0 +1,4 @@
+WARNING: The directory '/home/alexchan/.cache/pip' or its parent directory is not owned or is not writable by the current user. The cache has been disabled. Check the permissions and owner of that directory. If executing pip with sudo, you should use sudo's -H flag.
+Processing ./dist/linux-mmcv-cpu/mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
+Installing collected packages: mmcv
+Successfully installed mmcv-2.2.0

builds/linux-cpu/PUBLICATION_PIP_CHECK.txt

@@ -0,0 +1,2 @@
+WARNING: The directory '/home/alexchan/.cache/pip' or its parent directory is not owned or is not writable by the current user. The cache has been disabled. Check the permissions and owner of that directory. If executing pip with sudo, you should use sudo's -H flag.
+No broken requirements found.

builds/linux-cpu/PUBLICATION_SMOKE.txt

@@ -0,0 +1,10 @@
+/home/alexchan/anaconda3/envs/MAAP_cpu/lib/python3.10/site-packages/mmengine/optim/optimizer/zero_optimizer.py:11: DeprecationWarning: `TorchScript` support for functional optimizers is deprecated and will be removed in a future PyTorch release. Consider using the `torch.compile` optimizer instead.
+  from torch.distributed.optim import \
+/home/alexchan/.config/matplotlib is not a writable directory
+Matplotlib created a temporary cache directory at /tmp/matplotlib-v27c_bzn because there was an issue with the default path (/home/alexchan/.config/matplotlib); it is highly recommended to set the MPLCONFIGDIR environment variable to a writable directory, in particular to speed up the import of Matplotlib and to better support multiprocessing.
+Fontconfig error: No writable cache directories
+Fontconfig error: No writable cache directories
+/home/alexchan/anaconda3/envs/MAAP_cpu/lib/python3.10/site-packages/mmpose/datasets/datasets/utils.py:102: UserWarning: The metainfo config file "configs/_base_/datasets/ak.py" does not exist. A matched config file "/home/alexchan/anaconda3/envs/MAAP_cpu/lib/python3.10/site-packages/mmpose/.mim/configs/_base_/datasets/ak.py" will be used instead.
+  warnings.warn(
+Candidate: /tmp/maap3d-linux-cpu-review/site-packages/mmcv/__init__.py
+CPU OpenMMLab smoke passed: torch=2.4.0+cpu mmcv=2.2.0 mmpose=1.3.2 mmdet=3.3.0+maap3d.cfd5d3a9.1

builds/linux-cpu/README.md

@@ -0,0 +1,64 @@
+# Linux CPU MMCV wheel
+
+This builds a local MMCV 2.2.0 candidate for **CPython 3.10, Linux x86-64,
+and PyTorch 2.4.0+cpu**, matching the Linux CPU installer. Windows and macOS
+need separate builds. There are no CUDA operators in this artifact.
+
+From the repository root, with Docker available:
+
+```bash
+mkdir -p dist/linux-mmcv-cpu
+image=quay.io/pypa/manylinux2014_x86_64@sha256:55df5597db86b0a03d5f21a03b060271efffc78ef97123c60ee9c7ce0f10ec31
+docker run --rm --user "$(id -u):$(id -g)" --memory=7g --memory-swap=8g --cpus=2 \
+  -v "$PWD:/repo:ro" \
+  -v "$PWD/dist/linux-mmcv-cpu:/output" \
+  -e PYTHONDONTWRITEBYTECODE=1 \
+  -e MAAP3D_BUILD_IMAGE="$image" \
+  "$image" \
+  /opt/python/cp310-cp310/bin/python /repo/linux_mmcv_cpu/build_mmcv.py \
+  --output-dir /output
+```
+
+Use an empty output directory for each candidate. The build uses temporary
+environments inside the container and does not modify host Python packages.
+The MMCV source archive is checked against the installer's SHA-256 pin.
+Dependency pins and the actual compiler/Python versions are recorded in
+`BUILD_REPORT.json`, together with the pinned Docker image digest. Pip is
+bootstrapped to 25.0.1 to handle dependency metadata consistently.
+To repeat validation of an existing candidate, append `--verify-only` to the
+Python command. Its existing `BUILD_REPORT.json` must match the wheel; original
+build provenance is preserved when validation results are updated. Verification pins Pillow 9.5.0 so its binary wheel works on
+the container's older glibc baseline, along with NumPy 1.26.4.
+
+The wheel has a `1cpupt240` build tag. Auditwheel repairs its system-library
+dependencies against the manylinux2014/glibc 2.17 baseline. **PyTorch libraries
+remain external**, so this is not a standalone manylinux distribution: users
+must install the matching PyTorch CPU runtime and import Torch before MMCV
+operators. The manylinux tags do not promise compatibility with other Torch
+versions, CPU architectures, or musl/Alpine Linux.
+
+The script installs the wheel in a second virtual environment and runs the
+repository's CPU NMS, RoIAlign forward/backward, checkpoint-free HRNet, and
+Mask R-CNN feature checks, followed by `pip check`. This environment is still
+inside the build container; it does not establish a compiler-free installation
+of every dependency or validate checkpoint-backed MAAP3D workflows.
+
+Successful output includes the wheel, `SHA256SUMS`, `BUILD_REPORT.json`,
+`AUDITWHEEL.txt`, `SMOKE_TEST.txt`, `PIP_CHECK.txt`, and the verification
+environment's frozen requirements. Keep these files together.
+
+This command only builds and validates local files. It does not publish a
+wheel or change `install_package.py --cpu` to download it.
+
+## Candidate built on 2026-10-21
+
+- Wheel: `mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl`
+- SHA-256: `d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970`
+- CPU operator/model feature checks and `pip check` passed in the fresh
+  verification environment. The wheel's 537 RECORD hashes and license files
+  were checked separately; its compiled extension links to no CUDA libraries.
+- The first build used the distutils fallback because Ninja was absent from
+  PATH. The recipe now adds its build environment's bin directory to PATH.
+- Local artifacts and validation logs are under `dist/linux-mmcv-cpu/`, which
+  is ignored by Git. `INTEGRITY.json` records the independent file-hash and
+  ELF dependency checks.

builds/linux-cpu/SHA256SUMS

@@ -0,0 +1 @@
+d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970  mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

builds/linux-cpu/SMOKE_TEST.txt

@@ -0,0 +1,7 @@
+CPU OpenMMLab smoke passed: torch=2.4.0+cpu mmcv=2.2.0 mmpose=1.3.2 mmdet=3.3.0+maap3d.cfd5d3a9.1
+/tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages/mmengine/optim/optimizer/zero_optimizer.py:11: DeprecationWarning: `TorchScript` support for functional optimizers is deprecated and will be removed in a future PyTorch release. Consider using the `torch.compile` optimizer instead.
+  from torch.distributed.optim import \
+/tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages/mmengine/utils/package_utils.py:48: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
+  from pkg_resources import DistributionNotFound, get_distribution
+/tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages/mmpose/datasets/datasets/utils.py:102: UserWarning: The metainfo config file "configs/_base_/datasets/ak.py" does not exist. A matched config file "/tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages/mmpose/.mim/configs/_base_/datasets/ak.py" will be used instead.
+  warnings.warn(

builds/linux-cpu/VERIFICATION_LOG.txt

@@ -0,0 +1,359 @@
+Running: /opt/python/cp310-cp310/bin/python -m venv /tmp/maap3d-linux-cpu-hdnepvf2/verify
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip --disable-pip-version-check install pip==25.0.1
+Collecting pip==25.0.1
+  Downloading pip-25.0.1-py3-none-any.whl (1.8 MB)
+     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.8/1.8 MB 13.8 MB/s eta 0:00:00
+Installing collected packages: pip
+  Attempting uninstall: pip
+    Found existing installation: pip 23.0.1
+    Uninstalling pip-23.0.1:
+      Successfully uninstalled pip-23.0.1
+Successfully installed pip-25.0.1
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip --disable-pip-version-check install torch==2.4.0+cpu torchvision==0.19.0+cpu numpy==1.26.4 pillow==9.5.0 --index-url https://download.pytorch.org/whl/cpu --extra-index-url https://pypi.org/simple
+Looking in indexes: https://download.pytorch.org/whl/cpu, https://pypi.org/simple
+Collecting torch==2.4.0+cpu
+  Downloading https://download-r2.pytorch.org/whl/cpu/torch-2.4.0%2Bcpu-cp310-cp310-linux_x86_64.whl (195.0 MB)
+     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 195.0/195.0 MB 19.6 MB/s eta 0:00:00
+Collecting torchvision==0.19.0+cpu
+  Downloading https://download-r2.pytorch.org/whl/cpu/torchvision-0.19.0%2Bcpu-cp310-cp310-linux_x86_64.whl (1.6 MB)
+     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.6/1.6 MB 22.1 MB/s eta 0:00:00
+Collecting numpy==1.26.4
+  Downloading numpy-1.26.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (61 kB)
+Collecting pillow==9.5.0
+  Downloading Pillow-9.5.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (9.5 kB)
+Collecting filelock (from torch==2.4.0+cpu)
+  Downloading filelock-4.1.0-py3-none-any.whl.metadata (2.0 kB)
+Collecting typing-extensions>=4.8.0 (from torch==2.4.0+cpu)
+  Downloading typing_extensions-4.16.0-py3-none-any.whl.metadata (3.3 kB)
+Collecting sympy (from torch==2.4.0+cpu)
+  Downloading sympy-1.14.0-py3-none-any.whl.metadata (12 kB)
+Collecting networkx (from torch==2.4.0+cpu)
+  Downloading networkx-3.4.2-py3-none-any.whl.metadata (6.3 kB)
+Collecting jinja2 (from torch==2.4.0+cpu)
+  Downloading jinja2-3.1.6-py3-none-any.whl.metadata (2.9 kB)
+Collecting fsspec (from torch==2.4.0+cpu)
+  Downloading fsspec-2026.9.0-py3-none-any.whl.metadata (10 kB)
+Collecting MarkupSafe>=2.0 (from jinja2->torch==2.4.0+cpu)
+  Downloading markupsafe-3.0.4-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl.metadata (2.7 kB)
+Collecting mpmath<1.4,>=1.1.0 (from sympy->torch==2.4.0+cpu)
+  Downloading mpmath-1.3.0-py3-none-any.whl.metadata (8.6 kB)
+Downloading numpy-1.26.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (18.2 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 18.2/18.2 MB 30.6 MB/s eta 0:00:00
+Downloading Pillow-9.5.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.3 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 3.3/3.3 MB 27.4 MB/s eta 0:00:00
+Downloading typing_extensions-4.16.0-py3-none-any.whl (45 kB)
+Downloading filelock-4.1.0-py3-none-any.whl (133 kB)
+Downloading fsspec-2026.9.0-py3-none-any.whl (221 kB)
+Downloading jinja2-3.1.6-py3-none-any.whl (134 kB)
+Downloading networkx-3.4.2-py3-none-any.whl (1.7 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.7/1.7 MB 35.9 MB/s eta 0:00:00
+Downloading sympy-1.14.0-py3-none-any.whl (6.3 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6.3/6.3 MB 20.2 MB/s eta 0:00:00
+Downloading markupsafe-3.0.4-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl (20 kB)
+Downloading mpmath-1.3.0-py3-none-any.whl (536 kB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 536.2/536.2 kB 35.5 MB/s eta 0:00:00
+Installing collected packages: mpmath, typing-extensions, sympy, pillow, numpy, networkx, MarkupSafe, fsspec, filelock, jinja2, torch, torchvision
+Successfully installed MarkupSafe-3.0.4 filelock-4.1.0 fsspec-2026.9.0 jinja2-3.1.6 mpmath-1.3.0 networkx-3.4.2 numpy-1.26.4 pillow-9.5.0 sympy-1.14.0 torch-2.4.0+cpu torchvision-0.19.0+cpu typing-extensions-4.16.0
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip --disable-pip-version-check install numpy==1.26.4 scipy==1.10.1 setuptools==80.9.0 opencv-python==4.10.0.84 addict==2.4.0 mmengine==0.10.6 yapf==0.43.0
+Requirement already satisfied: numpy==1.26.4 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (1.26.4)
+Collecting scipy==1.10.1
+  Downloading scipy-1.10.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (58 kB)
+Collecting setuptools==80.9.0
+  Downloading setuptools-80.9.0-py3-none-any.whl.metadata (6.6 kB)
+Collecting opencv-python==4.10.0.84
+  Downloading opencv_python-4.10.0.84-cp37-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (20 kB)
+Collecting addict==2.4.0
+  Downloading addict-2.4.0-py3-none-any.whl.metadata (1.0 kB)
+Collecting mmengine==0.10.6
+  Downloading mmengine-0.10.6-py3-none-any.whl.metadata (20 kB)
+Collecting yapf==0.43.0
+  Downloading yapf-0.43.0-py3-none-any.whl.metadata (46 kB)
+Collecting matplotlib (from mmengine==0.10.6)
+  Downloading matplotlib-3.10.9-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (52 kB)
+Collecting pyyaml (from mmengine==0.10.6)
+  Downloading pyyaml-6.0.3-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl.metadata (2.4 kB)
+Collecting rich (from mmengine==0.10.6)
+  Downloading rich-15.0.0-py3-none-any.whl.metadata (18 kB)
+Collecting termcolor (from mmengine==0.10.6)
+  Downloading termcolor-3.3.0-py3-none-any.whl.metadata (6.5 kB)
+Collecting platformdirs>=3.5.1 (from yapf==0.43.0)
+  Downloading platformdirs-4.12.4-py3-none-any.whl.metadata (5.5 kB)
+Collecting tomli>=2.0.1 (from yapf==0.43.0)
+  Downloading tomli-2.5.0-py3-none-any.whl.metadata (10 kB)
+Collecting contourpy>=1.0.1 (from matplotlib->mmengine==0.10.6)
+  Downloading contourpy-1.3.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (5.5 kB)
+Collecting cycler>=0.10 (from matplotlib->mmengine==0.10.6)
+  Downloading cycler-0.12.1-py3-none-any.whl.metadata (3.8 kB)
+Collecting fonttools>=4.22.0 (from matplotlib->mmengine==0.10.6)
+  Downloading fonttools-4.65.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (125 kB)
+Collecting kiwisolver>=1.3.1 (from matplotlib->mmengine==0.10.6)
+  Downloading kiwisolver-1.5.1-cp310-cp310-manylinux_2_12_x86_64.manylinux2010_x86_64.whl.metadata (5.2 kB)
+Collecting packaging>=20.0 (from matplotlib->mmengine==0.10.6)
+  Downloading packaging-26.3-py3-none-any.whl.metadata (3.5 kB)
+Requirement already satisfied: pillow>=8 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmengine==0.10.6) (9.5.0)
+Collecting pyparsing>=3 (from matplotlib->mmengine==0.10.6)
+  Downloading pyparsing-3.3.3-py3-none-any.whl.metadata (5.9 kB)
+Collecting python-dateutil>=2.7 (from matplotlib->mmengine==0.10.6)
+  Downloading python_dateutil-2.9.0.post0-py2.py3-none-any.whl.metadata (8.4 kB)
+Collecting markdown-it-py>=2.2.0 (from rich->mmengine==0.10.6)
+  Downloading markdown_it_py-4.2.0-py3-none-any.whl.metadata (7.4 kB)
+Collecting pygments<3.0.0,>=2.13.0 (from rich->mmengine==0.10.6)
+  Downloading pygments-2.21.0-py3-none-any.whl.metadata (2.5 kB)
+Collecting mdurl~=0.1 (from markdown-it-py>=2.2.0->rich->mmengine==0.10.6)
+  Downloading mdurl-0.1.2-py3-none-any.whl.metadata (1.6 kB)
+Collecting six>=1.5 (from python-dateutil>=2.7->matplotlib->mmengine==0.10.6)
+  Downloading six-1.17.0-py2.py3-none-any.whl.metadata (1.7 kB)
+Downloading scipy-1.10.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (34.4 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 34.4/34.4 MB 18.9 MB/s eta 0:00:00
+Downloading setuptools-80.9.0-py3-none-any.whl (1.2 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.2/1.2 MB 21.2 MB/s eta 0:00:00
+Downloading opencv_python-4.10.0.84-cp37-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (62.5 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 62.5/62.5 MB 11.4 MB/s eta 0:00:00
+Downloading addict-2.4.0-py3-none-any.whl (3.8 kB)
+Downloading mmengine-0.10.6-py3-none-any.whl (452 kB)
+Downloading yapf-0.43.0-py3-none-any.whl (256 kB)
+Downloading platformdirs-4.12.4-py3-none-any.whl (32 kB)
+Downloading tomli-2.5.0-py3-none-any.whl (14 kB)
+Downloading matplotlib-3.10.9-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (8.8 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 8.8/8.8 MB 9.8 MB/s eta 0:00:00
+Downloading pyyaml-6.0.3-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl (770 kB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 770.3/770.3 kB 10.5 MB/s eta 0:00:00
+Downloading rich-15.0.0-py3-none-any.whl (310 kB)
+Downloading termcolor-3.3.0-py3-none-any.whl (7.7 kB)
+Downloading contourpy-1.3.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (325 kB)
+Downloading cycler-0.12.1-py3-none-any.whl (8.3 kB)
+Downloading fonttools-4.65.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (5.3 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━��━━━━━━━━━ 5.3/5.3 MB 11.3 MB/s eta 0:00:00
+Downloading kiwisolver-1.5.1-cp310-cp310-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (1.6 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.6/1.6 MB 11.6 MB/s eta 0:00:00
+Downloading markdown_it_py-4.2.0-py3-none-any.whl (91 kB)
+Downloading packaging-26.3-py3-none-any.whl (129 kB)
+Downloading pygments-2.21.0-py3-none-any.whl (1.3 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.3/1.3 MB 9.7 MB/s eta 0:00:00
+Downloading pyparsing-3.3.3-py3-none-any.whl (126 kB)
+Downloading python_dateutil-2.9.0.post0-py2.py3-none-any.whl (229 kB)
+Downloading mdurl-0.1.2-py3-none-any.whl (10.0 kB)
+Downloading six-1.17.0-py2.py3-none-any.whl (11 kB)
+Installing collected packages: addict, tomli, termcolor, six, setuptools, scipy, pyyaml, pyparsing, pygments, platformdirs, packaging, opencv-python, mdurl, kiwisolver, fonttools, cycler, contourpy, yapf, python-dateutil, markdown-it-py, rich, matplotlib, mmengine
+  Attempting uninstall: setuptools
+    Found existing installation: setuptools 79.0.1
+    Uninstalling setuptools-79.0.1:
+      Successfully uninstalled setuptools-79.0.1
+Successfully installed addict-2.4.0 contourpy-1.3.2 cycler-0.12.1 fonttools-4.65.0 kiwisolver-1.5.1 markdown-it-py-4.2.0 matplotlib-3.10.9 mdurl-0.1.2 mmengine-0.10.6 opencv-python-4.10.0.84 packaging-26.3 platformdirs-4.12.4 pygments-2.21.0 pyparsing-3.3.3 python-dateutil-2.9.0.post0 pyyaml-6.0.3 rich-15.0.0 scipy-1.10.1 setuptools-80.9.0 six-1.17.0 termcolor-3.3.0 tomli-2.5.0 yapf-0.43.0
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip --disable-pip-version-check install --no-deps /output/mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
+Processing /output/mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
+Installing collected packages: mmcv
+Successfully installed mmcv-2.2.0
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -c import install_package as i; i.install_patched_wheel('chumpy'); i.install_patched_wheel('mmdet')
+Processing /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/chumpy-0.71+maap3d.580566ea.1-py3-none-any.whl
+Requirement already satisfied: scipy>=0.13.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from chumpy==0.71+maap3d.580566ea.1) (1.10.1)
+Requirement already satisfied: six>=1.11.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from chumpy==0.71+maap3d.580566ea.1) (1.17.0)
+Requirement already satisfied: numpy<1.27.0,>=1.19.5 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from scipy>=0.13.0->chumpy==0.71+maap3d.580566ea.1) (1.26.4)
+Installing collected packages: chumpy
+Successfully installed chumpy-0.71+maap3d.580566ea.1
+
+[notice] A new release of pip is available: 25.0.1 -> 26.2.2
+[notice] To update, run: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip install --upgrade pip
+Processing /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/mmdet-3.3.0+maap3d.cfd5d3a9.1-py3-none-any.whl
+Requirement already satisfied: matplotlib in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmdet==3.3.0+maap3d.cfd5d3a9.1) (3.10.9)
+Requirement already satisfied: numpy in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmdet==3.3.0+maap3d.cfd5d3a9.1) (1.26.4)
+Collecting pycocotools (from mmdet==3.3.0+maap3d.cfd5d3a9.1)
+  Downloading pycocotools-2.0.11-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl.metadata (1.3 kB)
+Requirement already satisfied: scipy in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmdet==3.3.0+maap3d.cfd5d3a9.1) (1.10.1)
+Collecting shapely (from mmdet==3.3.0+maap3d.cfd5d3a9.1)
+  Downloading shapely-2.1.2-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (6.8 kB)
+Requirement already satisfied: six in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmdet==3.3.0+maap3d.cfd5d3a9.1) (1.17.0)
+Collecting terminaltables (from mmdet==3.3.0+maap3d.cfd5d3a9.1)
+  Downloading terminaltables-3.1.10-py2.py3-none-any.whl.metadata (3.5 kB)
+Collecting tqdm (from mmdet==3.3.0+maap3d.cfd5d3a9.1)
+  Downloading tqdm-4.70.1-py3-none-any.whl.metadata (57 kB)
+Requirement already satisfied: contourpy>=1.0.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (1.3.2)
+Requirement already satisfied: cycler>=0.10 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (0.12.1)
+Requirement already satisfied: fonttools>=4.22.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (4.65.0)
+Requirement already satisfied: kiwisolver>=1.3.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (1.5.1)
+Requirement already satisfied: packaging>=20.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (26.3)
+Requirement already satisfied: pillow>=8 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (9.5.0)
+Requirement already satisfied: pyparsing>=3 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (3.3.3)
+Requirement already satisfied: python-dateutil>=2.7 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmdet==3.3.0+maap3d.cfd5d3a9.1) (2.9.0.post0)
+Downloading pycocotools-2.0.11-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl (472 kB)
+Downloading shapely-2.1.2-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (3.1 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 3.1/3.1 MB 22.3 MB/s eta 0:00:00
+Downloading terminaltables-3.1.10-py2.py3-none-any.whl (15 kB)
+Downloading tqdm-4.70.1-py3-none-any.whl (80 kB)
+Installing collected packages: tqdm, terminaltables, shapely, pycocotools, mmdet
+Successfully installed mmdet-3.3.0+maap3d.cfd5d3a9.1 pycocotools-2.0.11 shapely-2.1.2 terminaltables-3.1.10 tqdm-4.70.1
+
+[notice] A new release of pip is available: 25.0.1 -> 26.2.2
+[notice] To update, run: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip install --upgrade pip
+Downloading pinned patched wheel: chumpy-0.71+maap3d.580566ea.1-py3-none-any.whl
+[OK] Downloaded and verified patched wheel: /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/chumpy-0.71+maap3d.580566ea.1-py3-none-any.whl
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip install /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/chumpy-0.71+maap3d.580566ea.1-py3-none-any.whl
+Downloading pinned patched wheel: mmdet-3.3.0+maap3d.cfd5d3a9.1-py3-none-any.whl
+[OK] Downloaded and verified patched wheel: /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/mmdet-3.3.0+maap3d.cfd5d3a9.1-py3-none-any.whl
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip install /tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/mmdet-3.3.0+maap3d.cfd5d3a9.1-py3-none-any.whl
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip --disable-pip-version-check install mmpose==1.3.2 mmpretrain==1.2.0 timm==1.0.17 numpy==1.26.4 scipy==1.10.1 opencv-python==4.10.0.84
+Collecting mmpose==1.3.2
+  Downloading mmpose-1.3.2-py2.py3-none-any.whl.metadata (29 kB)
+Collecting mmpretrain==1.2.0
+  Downloading mmpretrain-1.2.0-py2.py3-none-any.whl.metadata (20 kB)
+Collecting timm==1.0.17
+  Downloading timm-1.0.17-py3-none-any.whl.metadata (59 kB)
+Requirement already satisfied: numpy==1.26.4 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (1.26.4)
+Requirement already satisfied: scipy==1.10.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (1.10.1)
+Requirement already satisfied: opencv-python==4.10.0.84 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (4.10.0.84)
+Requirement already satisfied: chumpy in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmpose==1.3.2) (0.71+maap3d.580566ea.1)
+Collecting json-tricks (from mmpose==1.3.2)
+  Downloading json_tricks-3.17.3-py2.py3-none-any.whl.metadata (16 kB)
+Requirement already satisfied: matplotlib in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmpose==1.3.2) (3.10.9)
+Collecting munkres (from mmpose==1.3.2)
+  Downloading munkres-3.0.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl.metadata (13 kB)
+Requirement already satisfied: pillow in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmpose==1.3.2) (9.5.0)
+Requirement already satisfied: torchvision in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmpose==1.3.2) (0.19.0+cpu)
+Collecting xtcocotools>=1.12 (from mmpose==1.3.2)
+  Downloading xtcocotools-1.14.3-cp310-cp310-manylinux1_x86_64.whl.metadata (364 bytes)
+Collecting einops (from mmpretrain==1.2.0)
+  Downloading einops-0.8.2-py3-none-any.whl.metadata (13 kB)
+Collecting importlib-metadata (from mmpretrain==1.2.0)
+  Downloading importlib_metadata-9.0.1-py3-none-any.whl.metadata (4.5 kB)
+Collecting mat4py (from mmpretrain==1.2.0)
+  Downloading mat4py-0.6.0-py2.py3-none-any.whl.metadata (2.8 kB)
+Collecting modelindex (from mmpretrain==1.2.0)
+  Downloading modelindex-0.0.2-py3-none-any.whl.metadata (756 bytes)
+Requirement already satisfied: rich in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from mmpretrain==1.2.0) (15.0.0)
+Requirement already satisfied: torch in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from timm==1.0.17) (2.4.0+cpu)
+Requirement already satisfied: pyyaml in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from timm==1.0.17) (6.0.3)
+Collecting huggingface_hub (from timm==1.0.17)
+  Downloading huggingface_hub-2.2.0-py3-none-any.whl.metadata (16 kB)
+Collecting safetensors (from timm==1.0.17)
+  Downloading safetensors-0.8.0-cp310-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (4.2 kB)
+Requirement already satisfied: setuptools>=18.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from xtcocotools>=1.12->mmpose==1.3.2) (80.9.0)
+Collecting cython>=0.27.3 (from xtcocotools>=1.12->mmpose==1.3.2)
+  Downloading cython-3.3.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl.metadata (28 kB)
+Requirement already satisfied: contourpy>=1.0.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (1.3.2)
+Requirement already satisfied: cycler>=0.10 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (0.12.1)
+Requirement already satisfied: fonttools>=4.22.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (4.65.0)
+Requirement already satisfied: kiwisolver>=1.3.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (1.5.1)
+Requirement already satisfied: packaging>=20.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (26.3)
+Requirement already satisfied: pyparsing>=3 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (3.3.3)
+Requirement already satisfied: python-dateutil>=2.7 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from matplotlib->mmpose==1.3.2) (2.9.0.post0)
+Requirement already satisfied: six>=1.11.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from chumpy->mmpose==1.3.2) (1.17.0)
+Collecting click<9.0.0,>=8.4.2 (from huggingface_hub->timm==1.0.17)
+  Downloading click-8.5.0-py3-none-any.whl.metadata (2.6 kB)
+Requirement already satisfied: filelock>=3.10.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from huggingface_hub->timm==1.0.17) (4.1.0)
+Requirement already satisfied: fsspec>=2023.5.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from huggingface_hub->timm==1.0.17) (2026.9.0)
+Collecting hf-xet<2.0.0,>=1.6.0 (from huggingface_hub->timm==1.0.17)
+  Downloading hf_xet-1.7.0-cp38-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (4.9 kB)
+Collecting httpx2<3,>=2.0.0 (from huggingface_hub->timm==1.0.17)
+  Downloading httpx2-2.14.0-py3-none-any.whl.metadata (10 kB)
+Requirement already satisfied: tomli>=1.1.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from huggingface_hub->timm==1.0.17) (2.5.0)
+Requirement already satisfied: tqdm>=4.42.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from huggingface_hub->timm==1.0.17) (4.70.1)
+Requirement already satisfied: typing-extensions>=4.1.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from huggingface_hub->timm==1.0.17) (4.16.0)
+Collecting zipp>=3.20 (from importlib-metadata->mmpretrain==1.2.0)
+  Downloading zipp-4.1.1-py3-none-any.whl.metadata (3.6 kB)
+Collecting model-index (from modelindex->mmpretrain==1.2.0)
+  Downloading model_index-0.1.11-py3-none-any.whl.metadata (3.9 kB)
+Requirement already satisfied: markdown-it-py>=2.2.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from rich->mmpretrain==1.2.0) (4.2.0)
+Requirement already satisfied: pygments<3.0.0,>=2.13.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from rich->mmpretrain==1.2.0) (2.21.0)
+Requirement already satisfied: sympy in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from torch->timm==1.0.17) (1.14.0)
+Requirement already satisfied: networkx in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from torch->timm==1.0.17) (3.4.2)
+Requirement already satisfied: jinja2 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from torch->timm==1.0.17) (3.1.6)
+Collecting anyio>=4.10 (from httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading anyio-4.15.1-py3-none-any.whl.metadata (4.7 kB)
+Collecting httpcore2==2.14.0 (from httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading httpcore2-2.14.0-py3-none-any.whl.metadata (27 kB)
+Collecting idna>=3.18 (from httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading idna-3.20-py3-none-any.whl.metadata (7.2 kB)
+Collecting truststore>=0.10 (from httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading truststore-0.10.5-py3-none-any.whl.metadata (4.5 kB)
+Collecting h11>=0.16 (from httpcore2==2.14.0->httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading h11-0.16.0-py3-none-any.whl.metadata (8.3 kB)
+Requirement already satisfied: mdurl~=0.1 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from markdown-it-py>=2.2.0->rich->mmpretrain==1.2.0) (0.1.2)
+Requirement already satisfied: MarkupSafe>=2.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from jinja2->torch->timm==1.0.17) (3.0.4)
+Collecting markdown (from model-index->modelindex->mmpretrain==1.2.0)
+  Downloading markdown-3.10.3-py3-none-any.whl.metadata (5.1 kB)
+Collecting ordered-set (from model-index->modelindex->mmpretrain==1.2.0)
+  Downloading ordered_set-4.1.0-py3-none-any.whl.metadata (5.3 kB)
+Requirement already satisfied: mpmath<1.4,>=1.1.0 in /tmp/maap3d-linux-cpu-hdnepvf2/verify/lib/python3.10/site-packages (from sympy->torch->timm==1.0.17) (1.3.0)
+Collecting exceptiongroup>=1.0.2 (from anyio>=4.10->httpx2<3,>=2.0.0->huggingface_hub->timm==1.0.17)
+  Downloading exceptiongroup-1.3.1-py3-none-any.whl.metadata (6.7 kB)
+Downloading mmpose-1.3.2-py2.py3-none-any.whl (1.7 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.7/1.7 MB 14.9 MB/s eta 0:00:00
+Downloading mmpretrain-1.2.0-py2.py3-none-any.whl (1.6 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.6/1.6 MB 21.1 MB/s eta 0:00:00
+Downloading timm-1.0.17-py3-none-any.whl (2.5 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.5/2.5 MB 9.2 MB/s eta 0:00:00
+Downloading xtcocotools-1.14.3-cp310-cp310-manylinux1_x86_64.whl (436 kB)
+Downloading einops-0.8.2-py3-none-any.whl (65 kB)
+Downloading huggingface_hub-2.2.0-py3-none-any.whl (839 kB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 839.9/839.9 kB 5.6 MB/s eta 0:00:00
+Downloading importlib_metadata-9.0.1-py3-none-any.whl (27 kB)
+Downloading json_tricks-3.17.3-py2.py3-none-any.whl (27 kB)
+Downloading mat4py-0.6.0-py2.py3-none-any.whl (13 kB)
+Downloading modelindex-0.0.2-py3-none-any.whl (2.1 kB)
+Downloading munkres-3.0.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl (173 kB)
+Downloading safetensors-0.8.0-cp310-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (516 kB)
+Downloading click-8.5.0-py3-none-any.whl (125 kB)
+Downloading cython-3.3.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl (3.5 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 3.5/3.5 MB 5.0 MB/s eta 0:00:00
+Downloading hf_xet-1.7.0-cp38-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (4.2 MB)
+   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 4.2/4.2 MB 5.7 MB/s eta 0:00:00
+Downloading httpx2-2.14.0-py3-none-any.whl (96 kB)
+Downloading httpcore2-2.14.0-py3-none-any.whl (84 kB)
+Downloading zipp-4.1.1-py3-none-any.whl (10 kB)
+Downloading model_index-0.1.11-py3-none-any.whl (34 kB)
+Downloading anyio-4.15.1-py3-none-any.whl (132 kB)
+Downloading idna-3.20-py3-none-any.whl (69 kB)
+Downloading truststore-0.10.5-py3-none-any.whl (19 kB)
+Downloading markdown-3.10.3-py3-none-any.whl (110 kB)
+Downloading ordered_set-4.1.0-py3-none-any.whl (7.6 kB)
+Downloading exceptiongroup-1.3.1-py3-none-any.whl (16 kB)
+Downloading h11-0.16.0-py3-none-any.whl (37 kB)
+Installing collected packages: mat4py, json-tricks, zipp, truststore, safetensors, ordered-set, munkres, markdown, idna, hf-xet, h11, exceptiongroup, einops, cython, click, model-index, importlib-metadata, httpcore2, anyio, xtcocotools, modelindex, httpx2, mmpretrain, mmpose, huggingface_hub, timm
+Successfully installed anyio-4.15.1 click-8.5.0 cython-3.3.0 einops-0.8.2 exceptiongroup-1.3.1 h11-0.16.0 hf-xet-1.7.0 httpcore2-2.14.0 httpx2-2.14.0 huggingface_hub-2.2.0 idna-3.20 importlib-metadata-9.0.1 json-tricks-3.17.3 markdown-3.10.3 mat4py-0.6.0 mmpose-1.3.2 mmpretrain-1.2.0 model-index-0.1.11 modelindex-0.0.2 munkres-3.0.0 ordered-set-4.1.0 safetensors-0.8.0 timm-1.0.17 truststore-0.10.5 xtcocotools-1.14.3 zipp-4.1.1
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python /repo/openmmlab_cpu_smoke.py
+CPU OpenMMLab smoke passed: torch=2.4.0+cpu mmcv=2.2.0 mmpose=1.3.2 mmdet=3.3.0+maap3d.cfd5d3a9.1
+
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip check
+Running: /tmp/maap3d-linux-cpu-hdnepvf2/verify/bin/python -m pip freeze
+Running: gcc --version
+{
+  "wheel": "mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl",
+  "sha256": "d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970",
+  "size_bytes": 30511921,
+  "python": "3.10.22",
+  "architecture": "x86_64",
+  "glibc": [
+    "glibc",
+    "2.17"
+  ],
+  "compiler": "gcc (GCC) 10.2.1 20210130 (Red Hat 10.2.1-11)",
+  "torch": "2.4.0+cpu",
+  "numpy": "1.26.4",
+  "mmcv": "2.2.0",
+  "source": {
+    "filename": "mmcv-2.2.0.tar.gz",
+    "sha256": "ac479247e808d8802f89eadf04d4118de86bdfe81361ec5aed0cc1bf731c67c9",
+    "url": "https://files.pythonhosted.org/packages/e9/a2/57a733e7e84985a8a0e3101dfb8170fc9db92435c16afad253069ae3f9df/mmcv-2.2.0.tar.gz"
+  },
+  "extension": "mmcv/_ext.cpython-310-x86_64-linux-gnu.so",
+  "external_torch_libraries": [
+    "libc10.so",
+    "libtorch.so",
+    "libtorch_cpu.so",
+    "libtorch_python.so"
+  ],
+  "build_image": "quay.io/pypa/manylinux2014_x86_64@sha256:55df5597db86b0a03d5f21a03b060271efffc78ef97123c60ee9c7ce0f10ec31",
+  "validation": {
+    "cpu_nms": "passed",
+    "cpu_roi_align_forward_backward": "passed",
+    "mmpose_hrnet_features": "passed",
+    "mmdet_maskrcnn_features": "passed",
+    "pip_check": "No broken requirements found."
+  },
+  "limits": [
+    "CPython 3.10 and Linux x86-64 only; no CUDA operators",
+    "Requires the matching PyTorch CPU libraries; not a standalone manylinux wheel",
+    "No model checkpoints or complete MAAP3D installation tested",
+    "Local candidate; not published or wired into installers"
+  ]
+}

builds/linux-cpu/build_mmcv.py

@@ -0,0 +1,164 @@
+#!/usr/bin/env python3
+"""Build and verify the CPython 3.10/Linux x86-64/Torch 2.4.0 CPU wheel.
+
+Run inside the manylinux2014 container documented in README.md. This script
+uses disposable virtual environments and never publishes or edits installers.
+"""
+
+import argparse
+import hashlib
+import json
+import os
+from pathlib import Path
+import platform
+import subprocess
+import sys
+import tarfile
+import tempfile
+import zipfile
+
+
+REPO = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(REPO))
+import install_package as installer
+
+
+def run(command, **kwargs):
+    print("Running:", " ".join(map(str, command)), flush=True)
+    result = subprocess.run(list(map(str, command)), check=False, **kwargs)
+    if result.returncode and kwargs.get("capture_output"):
+        print(result.stdout, flush=True)
+        print(result.stderr, file=sys.stderr, flush=True)
+    result.check_returncode()
+    return result
+
+
+def pip(python, *arguments):
+    run([python, "-m", "pip", "--disable-pip-version-check", *arguments])
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--output-dir", type=Path, required=True)
+    parser.add_argument("--verify-only", action="store_true", help="Verify the existing candidate without rebuilding")
+    args = parser.parse_args()
+    if sys.platform != "linux" or platform.machine() != "x86_64":
+        raise RuntimeError("Build this wheel on Linux x86-64.")
+    if sys.version_info[:2] != (3, 10):
+        raise RuntimeError("This candidate targets CPython 3.10 only.")
+    if platform.libc_ver() != ("glibc", "2.17"):
+        raise RuntimeError("Use the documented manylinux2014/glibc 2.17 container.")
+    output = args.output_dir.resolve()
+    output.mkdir(parents=True, exist_ok=True)
+    if not args.verify_only and (list(output.glob("*.whl")) or (output / "BUILD_REPORT.json").exists()):
+        raise FileExistsError("Use an empty output directory for a new candidate.")
+
+    environment = os.environ.copy()
+    environment.update({
+        "PYTHONNOUSERSITE": "1", "PYTHONDONTWRITEBYTECODE": "1",
+        "PIP_NO_CACHE_DIR": "1", "MAX_JOBS": "2", "MMCV_WITH_OPS": "1",
+        "FORCE_CUDA": "0", "OMP_NUM_THREADS": "2", "MKL_NUM_THREADS": "2",
+    })
+    os.environ.update(environment)
+    with tempfile.TemporaryDirectory(prefix="maap3d-linux-cpu-") as temporary:
+        temporary = Path(temporary)
+        torch_libs = ["libc10.so", "libtorch.so", "libtorch_cpu.so", "libtorch_python.so"]
+        if not args.verify_only:
+            build_python = temporary / "build/bin/python"
+            run([sys.executable, "-m", "venv", temporary / "build"])
+            pip(build_python, "install", "pip==25.0.1")
+            pip(build_python, "install", "torch==2.4.0+cpu", "--index-url", installer.CPU_INDEX_URL)
+            pip(build_python, "install", "numpy==1.26.4", "setuptools==80.9.0",
+                "wheel==0.44.0", "ninja==1.11.1.4", "psutil==6.1.1", "auditwheel==6.4.2")
+            environment["PATH"] = str(build_python.parent) + os.pathsep + environment["PATH"]
+            run([build_python, "-c", "import torch; assert torch.__version__ == '2.4.0+cpu'; assert torch.version.cuda is None"])
+            source = Path(installer.download_mmcv_source(cache_dir=str(temporary / "source")))
+            with tarfile.open(source) as archive:
+                # The archive's exact SHA-256 is verified before extraction.
+                archive.extractall(temporary / "source-tree", filter="data")
+            source_root = temporary / "source-tree/mmcv-2.2.0"
+            raw = temporary / "raw"
+            run([build_python, "setup.py", "bdist_wheel", "--build-number", "1cpupt240",
+                 "--dist-dir", raw], cwd=source_root, env=environment)
+            wheels = list(raw.glob("*.whl"))
+            if len(wheels) != 1:
+                raise RuntimeError("Expected exactly one CPU wheel.")
+            auditwheel = temporary / "build/bin/auditwheel"
+            # Torch owns these ABI-matched libraries. Never bundle another copy.
+            repair = [auditwheel, "repair", "--plat", "manylinux2014_x86_64", "--wheel-dir", output]
+            for library in torch_libs:
+                repair.extend(["--exclude", library])
+            run([*repair, wheels[0]], env=environment)
+            wheel, = output.glob("*.whl")
+            audit = run([auditwheel, "show", wheel], capture_output=True, text=True)
+            (output / "AUDITWHEEL.txt").write_text(audit.stdout + audit.stderr)
+        wheel, = output.glob("*.whl")
+        with zipfile.ZipFile(wheel) as archive:
+            extension, = [name for name in archive.namelist()
+                          if name.startswith("mmcv/_ext") and name.endswith(".so")]
+            if not any("LICENSE" in name for name in archive.namelist()):
+                raise RuntimeError("The wheel must retain its upstream license.")
+            if archive.testzip() is not None:
+                raise RuntimeError("Corrupt wheel ZIP entry.")
+
+        report_path = output / "BUILD_REPORT.json"
+        if args.verify_only:
+            # Revalidation must retain the original compiler/source/image.
+            report = json.loads(report_path.read_text())
+            if report["wheel"] != wheel.name or report["sha256"] != hashlib.sha256(wheel.read_bytes()).hexdigest():
+                raise RuntimeError("Existing build report does not match this wheel.")
+        else:
+            compiler = run(["gcc", "--version"], capture_output=True, text=True).stdout.splitlines()[0]
+            report = {
+                "wheel": wheel.name, "sha256": hashlib.sha256(wheel.read_bytes()).hexdigest(),
+                "size_bytes": wheel.stat().st_size, "python": platform.python_version(),
+                "architecture": platform.machine(), "glibc": platform.libc_ver(),
+                "compiler": compiler, "torch": "2.4.0+cpu", "numpy": "1.26.4",
+                "mmcv": "2.2.0", "source": installer.MMCV_SOURCE_ARTIFACT,
+                "extension": extension, "external_torch_libraries": torch_libs,
+                "build_image": os.environ.get("MAAP3D_BUILD_IMAGE", "not recorded"),
+                "limits": ["CPython 3.10 and Linux x86-64 only; no CUDA operators",
+                           "Requires the matching PyTorch CPU libraries; not a standalone manylinux wheel",
+                           "No model checkpoints or complete MAAP3D installation tested",
+                           "Local candidate; not published or wired into installers"],
+            }
+            report["validation"] = {}
+            report_path.write_text(json.dumps(report, indent=2) + "\n")
+
+        # A new environment ensures source-tree/build dependencies cannot mask
+        # a missing wheel file or runtime requirement.
+        verify_python = temporary / "verify/bin/python"
+        run([sys.executable, "-m", "venv", temporary / "verify"])
+        pip(verify_python, "install", "pip==25.0.1")
+        pip(verify_python, "install", "torch==2.4.0+cpu", "torchvision==0.19.0+cpu",
+            "numpy==1.26.4", "pillow==9.5.0",
+            "--index-url", installer.CPU_INDEX_URL, "--extra-index-url", "https://pypi.org/simple")
+        pip(verify_python, "install", "numpy==1.26.4", "scipy==1.10.1",
+            "setuptools==80.9.0", "opencv-python==4.10.0.84",
+            *installer.MMCV_RUNTIME_PREREQUISITES)
+        pip(verify_python, "install", "--no-deps", str(wheel))
+        run([verify_python, "-c",
+             "import install_package as i; i.install_patched_wheel('chumpy'); i.install_patched_wheel('mmdet')"], cwd=REPO)
+        pip(verify_python, "install", "mmpose==1.3.2", "mmpretrain==1.2.0",
+            "timm==1.0.17", "numpy==1.26.4", "scipy==1.10.1",
+            "opencv-python==4.10.0.84")
+        smoke = run([verify_python, REPO / "openmmlab_cpu_smoke.py"], cwd=temporary,
+                    capture_output=True, text=True, env=environment)
+        (output / "SMOKE_TEST.txt").write_text(smoke.stdout + smoke.stderr)
+        print(smoke.stdout, flush=True)
+        checked = run([verify_python, "-m", "pip", "check"], capture_output=True, text=True)
+        (output / "PIP_CHECK.txt").write_text(checked.stdout + checked.stderr)
+        frozen = run([verify_python, "-m", "pip", "freeze"], capture_output=True, text=True)
+        (output / "verification-requirements.txt").write_text(frozen.stdout)
+        report["validation"] = {
+            "cpu_nms": "passed", "cpu_roi_align_forward_backward": "passed",
+            "mmpose_hrnet_features": "passed", "mmdet_maskrcnn_features": "passed",
+            "pip_check": checked.stdout.strip(),
+        }
+        (output / "BUILD_REPORT.json").write_text(json.dumps(report, indent=2) + "\n")
+        (output / "SHA256SUMS").write_text(report["sha256"] + "  " + wheel.name + "\n")
+        print(json.dumps(report, indent=2), flush=True)
+
+
+if __name__ == "__main__":
+    main()

builds/linux-cpu/openmmlab_cpu_smoke.py

@@ -0,0 +1,83 @@
+#!/usr/bin/env python3
+"""Exercise compiled MMCV ops and checkpoint-free OpenMMLab models on CPU."""
+
+import importlib.util
+import os
+
+from openmmlab_cuda_smoke import (
+    MMDET_CONFIG,
+    MMPOSE_CONFIG,
+    _assert_feature_tensors,
+    _disable_init_cfg,
+)
+
+
+def run_smoke():
+    import mmcv
+    import mmdet
+    import mmpose
+    import torch
+    from mmcv.ops import nms, roi_align
+    from mmengine.config import Config
+
+    if torch.version.cuda is not None or torch.version.hip is not None:
+        raise RuntimeError("The CPU smoke requires a CPU-only PyTorch build.")
+    if importlib.util.find_spec("mmcv._ext") is None:
+        raise RuntimeError("Compiled mmcv._ext is missing from the CPU profile.")
+
+    device = torch.device("cpu")
+    boxes = torch.tensor(
+        [[0.0, 0.0, 10.0, 10.0], [1.0, 1.0, 9.0, 9.0], [20.0, 20.0, 30.0, 30.0]]
+    )
+    scores = torch.tensor([0.9, 0.8, 0.7])
+    detections, kept = nms(boxes, scores, 0.5)
+    if kept.tolist() != [0, 2] or not torch.isfinite(detections).all():
+        raise RuntimeError("MMCV CPU NMS returned unexpected results.")
+
+    feature_map = torch.arange(256.0).reshape(1, 1, 16, 16).requires_grad_()
+    rois = torch.tensor([[0.0, 1.0, 1.0, 12.0, 12.0]])
+    pooled = roi_align(feature_map, rois, (4, 4), 1.0, 0, "avg", True)
+    pooled.sum().backward()
+    if not torch.isfinite(pooled).all() or feature_map.grad is None:
+        raise RuntimeError("MMCV CPU RoIAlign forward/backward failed.")
+    if not torch.isfinite(feature_map.grad).all():
+        raise RuntimeError("MMCV CPU RoIAlign returned invalid gradients.")
+
+    from mmpose.apis import init_model
+
+    pose_cfg = Config.fromfile(os.fspath(MMPOSE_CONFIG))
+    _disable_init_cfg(pose_cfg.model)
+    pose_model = init_model(pose_cfg, checkpoint=None, device="cpu").eval()
+    pose_size = tuple(pose_cfg.codec.input_size)
+    with torch.inference_mode():
+        pose_features = pose_model.extract_feat(
+            torch.zeros((1, 3, pose_size[1], pose_size[0]))
+        )
+    _assert_feature_tensors(pose_features, label="MMPose HRNet", device=device)
+    del pose_features, pose_model
+
+    from mmdet.registry import MODELS
+    from mmdet.utils import register_all_modules
+
+    register_all_modules(init_default_scope=True)
+    detection_cfg = Config.fromfile(os.fspath(MMDET_CONFIG))
+    _disable_init_cfg(detection_cfg.model)
+    detection_model = MODELS.build(detection_cfg.model).to(device).eval()
+    with torch.inference_mode():
+        detection_features = detection_model.extract_feat(
+            torch.zeros((1, 3, 256, 256))
+        )
+    _assert_feature_tensors(
+        detection_features, label="MMDetection Mask R-CNN", device=device
+    )
+    print(
+        "CPU OpenMMLab smoke passed:",
+        f"torch={torch.__version__}",
+        f"mmcv={mmcv.__version__}",
+        f"mmpose={mmpose.__version__}",
+        f"mmdet={mmdet.__version__}",
+    )
+
+
+if __name__ == "__main__":
+    run_smoke()

builds/linux-cpu/openmmlab_cuda_smoke.py

@@ -0,0 +1,253 @@
+#!/usr/bin/env python3
+"""Exercise compiled MMCV ops and MAAP3D OpenMMLab models on CUDA."""
+
+import argparse
+import importlib.resources
+import importlib.util
+import os
+from collections.abc import Mapping
+from pathlib import Path
+
+
+REPOSITORY_ROOT = Path(__file__).resolve().parent
+MMPOSE_CONFIG = REPOSITORY_ROOT / "MAAP3D/MMPoseConfigs/hrnet.py"
+MMDET_CONFIG = REPOSITORY_ROOT / "MAAP3D/MMDetConfigs/maskrcnn50.py"
+SAM3_TOKENIZER_ASSET = "assets/bpe_simple_vocab_16e6.txt.gz"
+
+
+def _disable_init_cfg(value):
+    """Recursively disable model initializers that could fetch checkpoints."""
+
+    if isinstance(value, Mapping):
+        for key in list(value):
+            if key == "init_cfg":
+                value[key] = None
+            else:
+                _disable_init_cfg(value[key])
+    elif isinstance(value, (list, tuple)):
+        for item in value:
+            _disable_init_cfg(item)
+
+
+def _assert_feature_tensors(features, *, label, device):
+    """Require a non-empty finite feature collection on the requested device."""
+
+    import torch
+
+    if isinstance(features, torch.Tensor):
+        tensors = [features]
+    elif isinstance(features, (list, tuple)):
+        tensors = [item for item in features if isinstance(item, torch.Tensor)]
+    else:
+        tensors = []
+    if not tensors:
+        raise RuntimeError(f"{label} returned no feature tensors.")
+    for tensor in tensors:
+        if tensor.device != device:
+            raise RuntimeError(
+                f"{label} returned a feature on {tensor.device}, expected {device}."
+            )
+        if not torch.isfinite(tensor).all():
+            raise RuntimeError(f"{label} returned non-finite feature values.")
+
+
+def _verify_sam3_import_surface():
+    """Verify SAM3 imports and its tokenizer asset without loading a checkpoint."""
+
+    from sam3.model_builder import build_sam3_video_predictor
+
+    if not callable(build_sam3_video_predictor):
+        raise RuntimeError("SAM3 video predictor builder is not callable.")
+    tokenizer_asset = importlib.resources.files("sam3").joinpath(
+        SAM3_TOKENIZER_ASSET
+    )
+    if not tokenizer_asset.is_file():
+        raise RuntimeError(f"SAM3 tokenizer asset is missing: {tokenizer_asset}")
+
+
+def _verify_sam3_triton_kernels(device):
+    """Compile and verify the SAM3 Triton kernels used by MAAP3D inference."""
+
+    import torch
+    from sam3.model.edt import edt_triton
+    from sam3.perflib.triton.connected_components import (
+        connected_components_triton,
+    )
+    from sam3.perflib.triton.nms import nms_triton
+
+    mask = torch.zeros((1, 4, 4), dtype=torch.bool, device=device)
+    mask[:, 1:3, 1:3] = True
+    distances = edt_triton(mask)
+    expected_distances = mask.to(dtype=torch.float32)
+    if not torch.equal(distances, expected_distances):
+        raise RuntimeError("SAM3 Triton EDT returned unexpected CUDA results.")
+
+    labels, component_sizes = connected_components_triton(mask.contiguous())
+    if labels.shape != mask.shape:
+        raise RuntimeError("SAM3 Triton connected-components changed the mask shape.")
+    if not torch.equal(
+        component_sizes[mask],
+        torch.full((4,), 4, dtype=torch.int32, device=device),
+    ):
+        raise RuntimeError(
+            "SAM3 Triton connected-components returned unexpected foreground sizes."
+        )
+    if not torch.equal(
+        component_sizes[~mask],
+        torch.zeros((12,), dtype=torch.int32, device=device),
+    ):
+        raise RuntimeError(
+            "SAM3 Triton connected-components returned unexpected background sizes."
+        )
+
+    ious = torch.tensor(
+        [[1.0, 0.8, 0.0], [0.8, 1.0, 0.0], [0.0, 0.0, 1.0]],
+        dtype=torch.float32,
+        device=device,
+    )
+    scores = torch.tensor([0.9, 0.8, 0.7], dtype=torch.float32, device=device)
+    kept_indices = nms_triton(ious, scores, 0.5)
+    if kept_indices.tolist() != [0, 2]:
+        raise RuntimeError("SAM3 Triton NMS returned unexpected CUDA results.")
+    torch.cuda.synchronize(device)
+
+
+def run_smoke(device="cuda:0", expected_cuda="12.8", verify_sam3_triton=False):
+    """Run CUDA operators plus checkpoint-free pose/detection feature passes."""
+
+    import mmcv
+    import mmdet
+    import mmpose
+    import torch
+    from mmcv.ops import nms, roi_align
+    from mmengine.config import Config
+
+    if not isinstance(device, str) or not device.strip():
+        raise ValueError("device must be a non-empty string.")
+    if not isinstance(expected_cuda, str) or not expected_cuda.strip():
+        raise ValueError("expected_cuda must be a non-empty string.")
+    if type(verify_sam3_triton) is not bool:
+        raise ValueError("verify_sam3_triton must be a boolean.")
+    requested_device = torch.device(device)
+    if requested_device.type != "cuda":
+        raise ValueError("The OpenMMLab smoke requires a cuda:* device.")
+    if torch.version.cuda != expected_cuda or torch.version.hip is not None:
+        raise RuntimeError(
+            "The OpenMMLab smoke requires PyTorch built for CUDA "
+            f"{expected_cuda}."
+        )
+    if not torch.cuda.is_available():
+        raise RuntimeError("PyTorch cannot access an NVIDIA CUDA GPU.")
+    if importlib.util.find_spec("mmcv._ext") is None:
+        raise RuntimeError(
+            "Compiled mmcv._ext is missing; mmcv-lite is not a full CUDA install."
+        )
+
+    boxes = torch.tensor(
+        [[0.0, 0.0, 10.0, 10.0], [1.0, 1.0, 9.0, 9.0], [20.0, 20.0, 30.0, 30.0]],
+        dtype=torch.float32,
+        device=requested_device,
+    )
+    scores = torch.tensor(
+        [0.9, 0.8, 0.7], dtype=torch.float32, device=requested_device
+    )
+    detections, kept_indices = nms(boxes, scores, iou_threshold=0.5)
+    if detections.device != requested_device or kept_indices.tolist() != [0, 2]:
+        raise RuntimeError("MMCV NMS returned unexpected CUDA results.")
+
+    feature_map = torch.arange(
+        256, dtype=torch.float32, device=requested_device
+    ).reshape(1, 1, 16, 16)
+    feature_map.requires_grad_(True)
+    rois = torch.tensor(
+        [[0.0, 1.0, 1.0, 12.0, 12.0]],
+        dtype=torch.float32,
+        device=requested_device,
+    )
+    pooled = roi_align(
+        feature_map,
+        rois,
+        (4, 4),
+        1.0,
+        0,
+        "avg",
+        True,
+    )
+    pooled.sum().backward()
+    if pooled.device != requested_device or not torch.isfinite(pooled).all():
+        raise RuntimeError("MMCV RoIAlign forward returned invalid CUDA output.")
+    if feature_map.grad is None or not torch.isfinite(feature_map.grad).all():
+        raise RuntimeError("MMCV RoIAlign backward returned invalid CUDA gradients.")
+    torch.cuda.synchronize(requested_device)
+
+    from mmpose.apis import init_model
+
+    pose_cfg = Config.fromfile(os.fspath(MMPOSE_CONFIG))
+    _disable_init_cfg(pose_cfg.model)
+    pose_model = init_model(pose_cfg, checkpoint=None, device=device)
+    pose_input_size = tuple(pose_cfg.codec.input_size)
+    pose_inputs = torch.zeros(
+        (1, 3, pose_input_size[1], pose_input_size[0]),
+        dtype=torch.float32,
+        device=requested_device,
+    )
+    with torch.inference_mode():
+        pose_features = pose_model.extract_feat(pose_inputs)
+    _assert_feature_tensors(
+        pose_features,
+        label="MMPose HRNet",
+        device=requested_device,
+    )
+    del pose_features, pose_inputs, pose_model
+    torch.cuda.empty_cache()
+
+    from mmdet.registry import MODELS
+    from mmdet.utils import register_all_modules
+
+    register_all_modules(init_default_scope=True)
+    detection_cfg = Config.fromfile(os.fspath(MMDET_CONFIG))
+    _disable_init_cfg(detection_cfg.model)
+    detection_model = MODELS.build(detection_cfg.model).to(requested_device).eval()
+    detection_inputs = torch.zeros(
+        (1, 3, 256, 256), dtype=torch.float32, device=requested_device
+    )
+    with torch.inference_mode():
+        detection_features = detection_model.extract_feat(detection_inputs)
+    _assert_feature_tensors(
+        detection_features,
+        label="MMDetection Mask R-CNN",
+        device=requested_device,
+    )
+    torch.cuda.synchronize(requested_device)
+
+    _verify_sam3_import_surface()
+    if verify_sam3_triton:
+        _verify_sam3_triton_kernels(requested_device)
+    print(
+        f"CUDA {expected_cuda} OpenMMLab smoke passed:",
+        f"device={torch.cuda.get_device_name(requested_device)}",
+        f"cuda={torch.version.cuda}",
+        f"torch={torch.__version__}",
+        f"mmcv={mmcv.__version__}",
+        f"mmpose={mmpose.__version__}",
+        f"mmdet={mmdet.__version__}",
+        "sam3=import-and-assets-ok",
+        f"sam3-triton={'verified' if verify_sam3_triton else 'not-requested'}",
+    )
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--device", default="cuda:0")
+    parser.add_argument("--expected-cuda", default="12.8")
+    parser.add_argument("--verify-sam3-triton", action="store_true")
+    args = parser.parse_args()
+    run_smoke(
+        device=args.device,
+        expected_cuda=args.expected_cuda,
+        verify_sam3_triton=args.verify_sam3_triton,
+    )
+
+
+if __name__ == "__main__":
+    main()

builds/linux-cpu/verification-requirements.txt

@@ -0,0 +1,67 @@
+addict==2.4.0
+anyio==4.15.1
+chumpy @ file:///tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/chumpy-0.71%2Bmaap3d.580566ea.1-py3-none-any.whl#sha256=d1fb98bf41865273eb4737a0391c2d2773c21eed15b6a0c747ec7190bd705974
+click==8.5.0
+contourpy==1.3.2
+cycler==0.12.1
+Cython==3.3.0
+einops==0.8.2
+exceptiongroup==1.3.1
+filelock==4.1.0
+fonttools==4.65.0
+fsspec==2026.9.0
+h11==0.16.0
+hf-xet==1.7.0
+httpcore2==2.14.0
+httpx2==2.14.0
+huggingface_hub==2.2.0
+idna==3.20
+importlib_metadata==9.0.1
+Jinja2==3.1.6
+json-tricks==3.17.3
+kiwisolver==1.5.1
+Markdown==3.10.3
+markdown-it-py==4.2.0
+MarkupSafe==3.0.4
+mat4py==0.6.0
+matplotlib==3.10.9
+mdurl==0.1.2
+mmcv @ file:///output/mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl#sha256=d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970
+mmdet @ file:///tmp/maap3d-patched-wheels/a7c5b8315db621aeb58755cfb8b1d3006a3c21f9/mmdet-3.3.0%2Bmaap3d.cfd5d3a9.1-py3-none-any.whl#sha256=ec27c030e40d0120e8f779308d6fcceed9f08eb4dd850eae89ce45fb0b63b579
+mmengine==0.10.6
+mmpose==1.3.2
+mmpretrain==1.2.0
+model-index==0.1.11
+modelindex==0.0.2
+mpmath==1.3.0
+munkres==3.0.0
+networkx==3.4.2
+numpy==1.26.4
+opencv-python==4.10.0.84
+ordered-set==4.1.0
+packaging==26.3
+Pillow==9.5.0
+platformdirs==4.12.4
+pycocotools==2.0.11
+Pygments==2.21.0
+pyparsing==3.3.3
+python-dateutil==2.9.0.post0
+PyYAML==6.0.3
+rich==15.0.0
+safetensors==0.8.0
+scipy==1.10.1
+shapely==2.1.2
+six==1.17.0
+sympy==1.14.0
+termcolor==3.3.0
+terminaltables==3.1.10
+timm==1.0.17
+tomli==2.5.0
+torch==2.4.0+cpu
+torchvision==0.19.0+cpu
+tqdm==4.70.1
+truststore==0.10.5
+typing_extensions==4.16.0
+xtcocotools==1.14.3
+yapf==0.43.0
+zipp==4.1.1

mmcv-2.2.0-1cpupt240-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d53f33353426cc2e17a170956c86c7920356b513b1eb4cf1dc80165096f0b970
+size 30511921