mirror of
https://github.com/pytorch/pytorch.git
synced 2025-10-20 21:14:14 +08:00
42f3d7e948
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129552 Approved by: https://github.com/malfet ghstack dependencies: #129451
45 lines
2.4 KiB
ReStructuredText
45 lines
2.4 KiB
ReStructuredText
.. _mps_environment_variables:
|
|
|
|
MPS Environment Variables
|
|
==========================
|
|
|
|
**PyTorch Environment Variables**
|
|
|
|
.. list-table::
|
|
:header-rows: 1
|
|
|
|
* - Variable
|
|
- Description
|
|
* - ``PYTORCH_DEBUG_MPS_ALLOCATOR``
|
|
- If set to ``1``, set allocator logging level to verbose.
|
|
* - ``PYTORCH_MPS_LOG_PROFILE_INFO``
|
|
- Set log options bitmask to ``MPSProfiler``. See ``LogOptions`` enum in `aten/src/ATen/mps/MPSProfiler.h` for available options.
|
|
* - ``PYTORCH_MPS_TRACE_SIGNPOSTS``
|
|
- Set profile and signpost bitmasks to ``MPSProfiler``. See ``ProfileOptions`` and ``SignpostTypes`` enums in `aten/src/ATen/mps/MPSProfiler.h` for available options.
|
|
* - ``PYTORCH_MPS_HIGH_WATERMARK_RATIO``
|
|
- High watermark ratio for MPS allocator. By default, it is set to 1.7.
|
|
* - ``PYTORCH_MPS_LOW_WATERMARK_RATIO``
|
|
- Low watermark ratio for MPS allocator. By default, it is set to 1.4 if the memory is unified and set to 1.0 if the memory is discrete.
|
|
* - ``PYTORCH_MPS_FAST_MATH``
|
|
- If set to ``1``, enable fast math for MPS metal kernels. See section 1.6.3 in https://developer.apple.com/metal/Metal-Shading-Language-Specification.pdf for precision implications.
|
|
* - ``PYTORCH_MPS_PREFER_METAL``
|
|
- If set to ``1``, force using metal kernels instead of using MPS Graph APIs. For now this is only used for matmul op.
|
|
* - ``PYTORCH_ENABLE_MPS_FALLBACK``
|
|
- If set to ``1``, full back operations to CPU when MPS does not support them.
|
|
|
|
.. note::
|
|
|
|
**high watermark ratio** is a hard limit for the total allowed allocations
|
|
|
|
- `0.0` : disables high watermark limit (may cause system failure if system-wide OOM occurs)
|
|
- `1.0` : recommended maximum allocation size (i.e., device.recommendedMaxWorkingSetSize)
|
|
- `>1.0`: allows limits beyond the device.recommendedMaxWorkingSetSize
|
|
|
|
e.g., value 0.95 means we allocate up to 95% of recommended maximum
|
|
allocation size; beyond that, the allocations would fail with OOM error.
|
|
|
|
**low watermark ratio** is a soft limit to attempt limiting memory allocations up to the lower watermark
|
|
level by garbage collection or committing command buffers more frequently (a.k.a, adaptive commit).
|
|
Value between 0 to m_high_watermark_ratio (setting 0.0 disables adaptive commit and garbage collection)
|
|
e.g., value 0.9 means we 'attempt' to limit allocations up to 90% of recommended maximum
|
|
allocation size. |