Add fused_partial_rope op #74577
Merged
Add fused_partial_rope op #74577
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
你的PR提交成功，感谢你对开源项目的贡献! |
f14d16f to
64ec214
Compare
64ec214 to
4710dc2
Compare
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #74577 +/- ##
==========================================
Coverage ? 97.29%
==========================================
Files ? 3
Lines ? 37
Branches ? 0
==========================================
Hits ? 36
Misses ? 1
Partials ? 0 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
zhangbo9674
approved these changes
Aug 15, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR Category
Operator Mechanism
PR Types
New features
Description
新增
paddle.incubate.nn.functional.fused_partial_rope(x, cos, sin)算子该算子是DeepseekV3使用的一种“部分RoPE“算子，即对于输入 x[bs, seq_len, num_heads, head_dim]，首先将 head_dim 拆分成 nope_head_dim + pe_head_dim，然后只在 pe_head_dim 上进行RoPE，最后合并回来
注意到，该算子只有一个输入x，因此 q 和 k 需要分别处理；这是模仿DSV3的timeline实现的，因为它的 q 比 k 大很多，分开处理更好实现
PaddleNLP修改已合入：PaddlePaddle/PaddleNLP#10942
（用了分支的形式，旧版的Paddle还能跑老的分支，新版Paddle就自动使用融合算子）
Pcard-85711