# PyTorch 2.4 Release Branch cut for PyTorch core is completed **URL:** \ **Category:** Release Announcements\ **Created:** [June 12, 2024, 11:26am UTC](https://dev-discuss.pytorch.org/t/pytorch-2-4-release-branch-cut-for-pytorch-core-is-completed/2152 "2024-06-12T11:26:53Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![atalman](https://yyz2.discourse-cdn.com/flex036/user_avatar/dev-discuss.pytorch.org/atalman/32/1857_2.png) [@atalman](https://dev-discuss.pytorch.org/u/atalman)\ **Post date:** [June 12, 2024, 11:26am UTC](https://dev-discuss.pytorch.org/t/pytorch-2-4-release-branch-cut-for-pytorch-core-is-completed/2152/1 "2024-06-12T11:26:53Z")
Release branch cut has been completed. The core binaries should now be available for download and testing on pytorch test channel. Here is the link to [HUD](https://hud.pytorch.org/hud/pytorch/pytorch/release%2F2.4/1?per_page=50) Release tracker issue for cherry-pick submission is available [here](https://github.com/pytorch/pytorch/issues/128436) Reminder of key dates: - M3.1: Release branch cut (6/10/24) - COMPLETED - M3.2: Release first RC1 Binary for PyTorch Core (6/11/24) - COMPLETED - M3.3: Domain libraries cut RC Branch (6/13/24) - M4: Release branch finalized, Announce final launch date, Feature classifications published (week of 7/1/24) - M5: External-Facing Content Finalized (7/19/24) - M6: Release Day (7/24/24) **Important Note regarding Triton package** : PyTorch 2.4 RC1 was built against following commit of Triton: [[backend][amd] Reorder and adjust optimization passes](https://github.com/openai/triton/commit/45fff310c891f5a92d55445adf8cc9d29df5841e) . However future PyTorch 2.4 RCs will be built against dedicated triton release branch. If you have any questions please comment below or reach out to us on this thread. Following are instructions on how to download different versions of RC1 for testing. CONDA CPU: Windows/LInux: conda install pytorch=2.4.0 cpuonly -c pytorch-test MacOS: conda install pytorch=2.4.0 -c pytorch-test CUDA 11.8, 12.1, 12.4 Windows/Linux: conda install pytorch=2.4.0 pytorch-cuda=11.8 -c pytorch-test -c nvidia conda install pytorch=2.4.0 pytorch-cuda=12.1 -c pytorch-test -c nvidia conda install pytorch=2.4.0 pytorch-cuda=12.4 -c pytorch-test -c nvidia PIP CPU: Windows/Linux/MacOS: pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/cpu](https://download.pytorch.org/whl/test/cpu) PIP CUDA 11.8, 12.1, 12.4 Windows/Linux: pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/cu118](https://download.pytorch.org/whl/test/cu118) pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/cu121](https://download.pytorch.org/whl/test/cu121) pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/cu124](https://download.pytorch.org/whl/test/cu124) PIP ROCM 6.0 pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/rocm6.0/](https://download.pytorch.org/whl/test/rocm6.0/) PIP ROCM 6.1 pip3 install torch==2.4.0 --index-url [https://download.pytorch.org/whl/test/rocm6.1/](https://download.pytorch.org/whl/test/rocm6.1/) Libtorch: CPU Linux: (Pre-cxx11 ABI): [https://download.pytorch.org/libtorch/test/cpu/libtorch-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cpu/libtorch-shared-with-deps-latest.zip) (cxx11 ABI): [https://download.pytorch.org/libtorch/test/cpu/libtorch-cxx11-abi-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cpu/libtorch-cxx11-abi-shared-with-deps-latest.zip) CPU Windows: Download here (Release version): [https://download.pytorch.org/libtorch/test/cpu/libtorch-win-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cpu/libtorch-win-shared-with-deps-latest.zip) Download here (Debug version): [https://download.pytorch.org/libtorch/test/cpu/libtorch-win-shared-with-deps-debug-latest.zip](https://download.pytorch.org/libtorch/test/cpu/libtorch-win-shared-with-deps-debug-latest.zip) CPU MacOS: [https://download.pytorch.org/libtorch/test/cpu/libtorch-macos-latest.zip](https://download.pytorch.org/libtorch/test/cpu/libtorch-macos-latest.zip) CUDA 11.8, 12.1, 12.4 Linux (Pre-cxx11 ABI): [https://download.pytorch.org/libtorch/test/cu118/libtorch-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu118/libtorch-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu121/libtorch-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu121/libtorch-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu124/libtorch-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu124/libtorch-shared-with-deps-latest.zip) (cxx11 ABI): [https://download.pytorch.org/libtorch/test/cu118/libtorch-cxx11-abi-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu118/libtorch-cxx11-abi-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu121/libtorch-cxx11-abi-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu121/libtorch-cxx11-abi-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu124/libtorch-cxx11-abi-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu124/libtorch-cxx11-abi-shared-with-deps-latest.zip) Windows CUDA 11.8, 12.1, 12.4 (Release version): [https://download.pytorch.org/libtorch/test/cu118/libtorch-win-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu118/libtorch-win-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu121/libtorch-win-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu121/libtorch-win-shared-with-deps-latest.zip) [https://download.pytorch.org/libtorch/test/cu124/libtorch-win-shared-with-deps-latest.zip](https://download.pytorch.org/libtorch/test/cu124/libtorch-win-shared-with-deps-latest.zip) (Debug version): [https://download.pytorch.org/libtorch/test/cu118/libtorch-win-shared-with-deps-debug-latest.zip](https://download.pytorch.org/libtorch/test/cu118/libtorch-win-shared-with-deps-debug-latest.zip) [https://download.pytorch.org/libtorch/test/cu121/libtorch-win-shared-with-deps-debug-latest.zip](https://download.pytorch.org/libtorch/test/cu121/libtorch-win-shared-with-deps-debug-latest.zip) [https://download.pytorch.org/libtorch/test/cu124/libtorch-win-shared-with-deps-debug-latest.zip](https://download.pytorch.org/libtorch/test/cu124/libtorch-win-shared-with-deps-debug-latest.zip) If you have any questions please comment below or reach out to us on this thread. Cheers, Team PyTorch #oneteam ---
**Author:** ![ashim-mahara](https://yyz2.discourse-cdn.com/flex036/user_avatar/dev-discuss.pytorch.org/ashim-mahara/32/2059_2.png) [@ashim-mahara](https://dev-discuss.pytorch.org/u/ashim-mahara)\ **Post date:** [July 10, 2024, 5:56am UTC](https://dev-discuss.pytorch.org/t/pytorch-2-4-release-branch-cut-for-pytorch-core-is-completed/2152/2 "2024-07-10T05:56:25Z")
Would the 2.4 release contain builds for aarch64 (grace-hopper systems) ? ---
**Author:** ![atalman](https://yyz2.discourse-cdn.com/flex036/user_avatar/dev-discuss.pytorch.org/atalman/32/1857_2.png) [@atalman](https://dev-discuss.pytorch.org/u/atalman)\ **Post date:** [July 11, 2024, 5:47pm UTC](https://dev-discuss.pytorch.org/t/pytorch-2-4-release-branch-cut-for-pytorch-core-is-completed/2152/3 "2024-07-11T17:47:20Z")
Hi @ashim-mahara Yes. These are prototype binaries. You should be able to install them using: `pip3 install torch --index-url https://download.pytorch.org/whl/test/cu124` Please note: Only torch binary was built for aarch64 (grace-hopper systems) torchvision and torchaudio are not availbale yet. ---
**Author:** ![ashim-mahara](https://yyz2.discourse-cdn.com/flex036/user_avatar/dev-discuss.pytorch.org/ashim-mahara/32/2059_2.png) [@ashim-mahara](https://dev-discuss.pytorch.org/u/ashim-mahara)\ **Post date:** [July 17, 2024, 12:53pm UTC](https://dev-discuss.pytorch.org/t/pytorch-2-4-release-branch-cut-for-pytorch-core-is-completed/2152/4 "2024-07-17T12:53:44Z")
Happy to report that it is working without errors on GH200 systems. - Tested single-node training - Tested multi-node training I am not able to use Infiniband but I think it is more related to my own network and configurations rather than torch.