trunk/40e238cee98065a2526d8420affade6be07912d7: [XPU] Fix accuracy issue in addmm for bf16/f16 dtypes (#174864)
- PyTorch 近 90 天出现 296 次
- PyTorch 近 90 天第 293 次发版
- 上一次：同一天稍早 · trunk/74ee2f44746e2868616296c50bed17b4e9b59376: [BE][Ez]: Enable CPP20 heterogenous lookups in STL collections (#195372)
发生了什么
Fixes intel/torch-xpu-ops#2837 Fixes precision loss in XPU addmm and baddbmm for bf16/f16 dtypes with non-trivial alpha/beta. The oneDNN 3-step post-op chain rounds intermediates to reduced precision at each stage. The fix pre-copies self into result and uses post_sum which accumulates in oneDNN's internal f32 accumulator, matching CPU/CUDA behavior. Pull Request resolved: #174864 Approved by: https://github.com/Eik…
摘要按规则整理自下方来源原文