Skip to content

v1.2.2

Choose a tag to compare

@LoserCheems LoserCheems released this 05 Nov 08:10
· 74 commits to main since this release

What's Changed

  • [FEATURE SUPPORT] Robust dBias accumulation for seqlen_q_bias == 1 by @LoserCheems in #194
  • [FEATURE SUPPORT] Centralize dynamic mask creation for FDMA by @LoserCheems in #197
  • Update documentation to use mask utility in examples by @LoserCheems in #198
  • Fix attention bias calculation and dbias handling by @LoserCheems in #199
  • Add block-wise smoothing to attention mask by @LoserCheems in #201
  • [FEATURE SUPPORT] Move scaling out of streaming loops, bias-initialized acc_s, and fix dQ double-scaling by @LoserCheems in #203

Full Changelog: v1.2.1...v1.2.2