Code

[feature] refactor colo attention#5462

Code

Merged

ver217 merged 7 commits intohpcaitech:feature/colo-attentionhpcaitech/ColossalAI:feature/colo-attentionfrom ver217:feature/colo-attnver217/ColossalAI:feature/colo-attnCopy head branch name to clipboardMar 15, 2024

Commits

Commits on Mar 13, 2024

• [extension] update api

  ver217

  committed
  07995a7

  Copy full SHA for 07995a7

• [feature] add colo attention

  ver217

  committed
  506b9cc

  Copy full SHA for 506b9cc

• [feature] update sdpa

  ver217

  committed
  28ea4e2

  Copy full SHA for 28ea4e2

• [feature] update npu attention

  ver217

  committed
  f8f4a21

  Copy full SHA for f8f4a21

• [feature] update flash-attn

  ver217

  committed
  dc275be

  Copy full SHA for dc275be

Commits on Mar 14, 2024

• [test] add flash attn test

  ver217

  committed
  112e182

  Copy full SHA for 112e182

• [test] update flash attn test

  ver217

  committed
  2e94583

  Copy full SHA for 2e94583