Sarthak Tandon 66ea76ec44 [ROCm][tunableop] Improvements to tunableop Numerical Check (#163079) ...

Modified the flag PYTORCH_TUNABLEOP_NUMERICAL_CHECK, so that it accepts the numerical tolerances in the format atol_rtol as compared to the previous 0 and 1. Retains previous functionality with default values as well.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163079
Approved by: https://github.com/naromero77amd, https://github.com/jeffdaily

2025-10-15 22:26:47 +00:00

..

shared

Revert "Add magic TORCH_MAKE_PYBIND_ENUM_FASTER macro (#163527)"

2025-10-02 15:42:42 +00:00

comm.cpp

Turn static inline into static function (#139843)

2024-11-07 23:58:18 +00:00

comm.h

[1/N] Change #include <c10/util/Optional.h> to #include <optional> (#128301)

2024-07-08 07:03:53 +00:00

CUDAPluggableAllocator.cpp

[PyTorch CCA] Add an API to get expandable segment sizes (#163771)

2025-10-01 02:16:58 +00:00

CUDAPluggableAllocator.h

[PyTorch CCA] Add an API to get expandable segment sizes (#163771)

2025-10-01 02:16:58 +00:00

device_set.h

Revert "Increased compile time max GPUs to 512. Switched to int16_t DeviceIndex. (#119639)"

2024-02-28 18:57:09 +00:00

Event.cpp

[2/N] Fix cppcoreguidelines-init-variables suppression (#146237)

2025-06-19 23:26:42 +00:00

Event.h

Make device-specific event inherits from torch.Event (#134845)

2024-10-01 06:28:41 +00:00

GdsFile.cpp

[1/N] Fix clang-tidy readability checks (#164561)

2025-10-04 09:40:38 +00:00

GdsFile.h

Fix initGdsBindings declaration (#152277)

2025-04-27 17:04:56 +00:00

Graph.cpp

[CUDAGraph] Add getter for cuda graph exec (#161294)

2025-08-25 20:57:37 +00:00

memory_snapshot.cpp

[Code Clean] Replace std::runtime_error with TORCH_CHECK (#163437)

2025-10-12 01:23:02 +00:00

memory_snapshot.h

[Memory Snapshot] Add Flag to Toggle Global and Local Callbacks for Annotations (#154932)

2025-06-04 23:15:19 +00:00

MemPool.cpp

[nccl symm mem] don't use arg for mempool, correctly use symmetric registration in hooks (#161238)

2025-08-25 03:09:32 +00:00

Module.cpp

[ROCm][tunableop] Improvements to tunableop Numerical Check (#163079)

2025-10-15 22:26:47 +00:00

Module.h

[11/N] Fix extra warnings brought by clang-tidy-17 (#139599)

2024-11-04 23:57:41 +00:00

nccl.cpp

[Code Clean] Replace std::runtime_error with TORCH_CHECK (#163437)

2025-10-12 01:23:02 +00:00

nccl.h

[9/N] Fix extra warnings brought by clang-tidy-17 (#139286)

2024-10-31 05:20:31 +00:00

python_comm.cpp

[8/N] Fix extra warnings brought by clang-tidy-17 (#139151)

2024-10-30 14:20:08 +00:00

python_comm.h

[8/N] Fix extra warnings brought by clang-tidy-17 (#139151)

2024-10-30 14:20:08 +00:00

python_nccl.cpp

Turn some const strings into constexpr in C++ code (#165203)

2025-10-13 20:25:20 +00:00

python_nccl.h

[pytorch] Add binding to get nccl version suffix (#112884)

2023-11-08 02:51:22 +00:00

Stream.cpp

[8/N] Fix extra warnings brought by clang-tidy-17 (#139151)

2024-10-30 14:20:08 +00:00

Stream.h

[Clang-tidy header][17/N] Apply clang-tidy on headers in torch/csrc/cuda (#117829)

2024-01-26 13:33:24 +00:00

THCP.h

Make torch/csrc/utils.h to be device-agnostic (#152521)

2025-05-04 07:15:11 +00:00

utils.cpp

[Code Clean] Replace std::runtime_error with TORCH_CHECK (#163437)

2025-10-12 01:23:02 +00:00

utils.h

Make torch/csrc/utils.h to be device-agnostic (#152521)

2025-05-04 07:15:11 +00:00