Denis Vieriu

[MPS - DRAFT] Add support for slice_scatter; enable index_put

5

Summary of changes: - support for scatter slice - enable index put With whole model delegation, I am seeing following crash in llama2: ``` in _verify_exported_program_signature raise SpecViolationError( torch._export.verifier.SpecViolationError: Buffer...

[MPS] Native nonzero implementation

2

Fixes https://github.com/pytorch/pytorch/issues/124850 Replace previous MPSGraph nonzero construction with native nonzero op. For older OSes, fallback to CPU (previous implementation was not reliable and was comparable to CPU in speed). cc...

Add support for cumprod

2

- Add support for cumprod. - Extend cumsum testcase to test cumprod.

Execute convolution in NCHW if the suggested mem format is NHWC but the actual mem layout is NCHW

1

Since **NHWC** is represented as a view operation in PyTorch, we can execute the convolution ops directly in NCHW if the **suggested memory format** is NHWC but the **actual memory...

[MPS] Add support for Int4 groupwise quantization

5

Add support for MPS Int4 per channel group-wise quantization through MPSGraph. --- Testing: **AOT export** ``` python -m examples.models.llama2.export_llama --checkpoint /Volumes/Source/weights/llama2/llama2-7b/llama-2-7b/consolidated.00.pth --params /Volumes/Source/weights/llama2/llama2-7b/llama-2-7b/params.json -kv --use_sdpa_with_kv_cache --mps -d fp32 --disable_dynamic_shape -qmode...