czhu-cohere 2c2b140ae8

[quantization] use channel scales for w4a8 + misc fixes (#23570) ...

Signed-off-by: czhu-cohere <conway.zhu@cohere.com>

2025-08-26 18:23:23 -07:00

..

__init__.py

[kernel] Support W4A8 on Hopper (#23198)

2025-08-24 06:18:04 +00:00

allspark.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

bitblas.py

[Bugfix] Add proper comparison for package versions (#22314)

2025-08-06 20:31:03 -07:00

conch.py

[Kernel] Add Conch backend for mixed-precision linear layer (#19818)

2025-07-09 13:17:55 -07:00

cutlass.py

[quantization] use channel scales for w4a8 + misc fixes (#23570)

2025-08-26 18:23:23 -07:00

dynamic_4bit.py

[Feat]: Add support for Dynamic Quant 4 bit CPU kleidiai kernels (#17112)

2025-07-28 20:55:15 +00:00

exllama.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

machete.py

[Bugfix] Fix Machete zero point issue for GPTQ models on SM90 (#21066)

2025-07-16 19:54:45 -07:00

marlin.py

[Kernel] [Quantization] Add MXFP4 and bias support for marlin kernel (#22428)

2025-08-14 11:23:22 -07:00

MPLinearKernel.py

[quantization] use channel scales for w4a8 + misc fixes (#23570)

2025-08-26 18:23:23 -07:00