-
Zhengju Tang authored
* [MXFP4] Dequantize FP4 kernel example, MX scale todo * [BugFix] Fix the bug of fp4&fp16 exponential bias * [MXFP4] Add group scale factor for BF16xMXFP4 gemm * [Lint] * [Test] Add test script for BF16xMXFP4 gemm * [Lint] * [BugFix] Fix the shape of scale tensor * Update example_dequant_gemm_fp4_hopper.py --------- Co-authored-by:
LeiWang1999 <leiwang1999@outlook.com> Co-authored-by:
Lei Wang <34334180+LeiWang1999@users.noreply.github.com>
569b0127