Yejing Lai d79bd930d6 Add cpu accelerator fp16 dtype support (#7207) ...

Add cpu accelerator fp16 dtype support

---------

Signed-off-by: Lai, Yejing <yejing.lai@intel.com>
Co-authored-by: Logan Adams <114770087+loadams@users.noreply.github.com>

2025-04-21 19:21:37 +00:00

..

__init__.py

[NPU] Add NPU support for unit test (#4569)

2023-11-13 20:36:12 +00:00

abstract_accelerator.py

assumption of torch.initial_seed function accepting seed arg in DeepSpeedAccelerator abstract class is incorrect (#5569)

2024-06-12 09:32:17 -07:00

cpu_accelerator.py

Add cpu accelerator fp16 dtype support (#7207)

2025-04-21 19:21:37 +00:00

cuda_accelerator.py

Fixed Windows inference build. (#5609)

2024-06-24 13:39:18 -07:00

hpu_accelerator.py

HPU accelerator memory mapping is broken because of torch fill uninit memory (#7209)

2025-04-08 18:38:06 +00:00

mlu_accelerator.py

[Accelerator] Cambricon MLU support (#6472)

2024-09-26 13:10:52 +00:00

mps_accelerator.py

assumption of torch.initial_seed function accepting seed arg in DeepSpeedAccelerator abstract class is incorrect (#5569)

2024-06-12 09:32:17 -07:00

npu_accelerator.py

assumption of torch.initial_seed function accepting seed arg in DeepSpeedAccelerator abstract class is incorrect (#5569)

2024-06-12 09:32:17 -07:00

real_accelerator.py

Tecorigin sdaa accelerator (#6903)

2025-02-20 20:46:21 +00:00

sdaa_accelerator.py

Tecorigin sdaa accelerator (#6903)

2025-02-20 20:46:21 +00:00

xpu_accelerator.py

[XPU] host timer check version from Torch 2.5 to Torch 2.6 (#6633)

2024-10-22 06:53:15 +00:00