Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Prim][PIR] Support composite rules of Llama ops #58018

Merged
merged 11 commits into from
Oct 13, 2023
Merged

[Prim][PIR] Support composite rules of Llama ops #58018

merged 11 commits into from
Oct 13, 2023

Conversation

cyber-pioneer
Copy link
Contributor

@cyber-pioneer cyber-pioneer commented Oct 11, 2023 •
edited
Loading

PR types

New features

PR changes

Others

Description

Pcard-66975

Llama 模型依赖的组合算子规则：p1

api 前向拆解 反向拆解 备注
def erf() python/paddle/tensor/math.py   p1 todo: float64 rtol > 1e-8
def expand() python/paddle/tensor/manipulation.py   p1  
def log() python/paddle/tensor/math.py   p1  
def sqrt() python/paddle/tensor/ops.py p1 p1  
def silu() python/paddle/nn/functional/activation.py p1 p1  
def softmax() python/paddle/nn/functional/activation.py p1 p1 todo：TestSoftmaxOp_ZeroDim
def full_like() python/paddle/tensor/creation.py p1    
def stack() python/paddle/tensor/manipulation.py p1    
def squeeze() python/paddle/tensor/manipulation.py p1    
def unsqueeze() python/paddle/tensor/manipulation.py p1    
def exp() python/paddle/tensor/ops.py   p1  
def gather_nd() python/paddle/tensor/manipulation.py   p1  
def pad() python/paddle/nn/functional/common.py   p1  
def slice() python/paddle/tensor/manipulation.py   p1  
def tile() python/paddle/tensor/manipulation.py   p1  



@paddle-bot
Copy link

paddle-bot bot commented Oct 11, 2023

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

@paddle-bot
Copy link

paddle-bot bot commented Oct 11, 2023

❌ The PR is not created using PR's template. You can refer to this Demo.
Please use PR's template, it helps save our maintainers' time so that more developers get helped.

@cyber-pioneer cyber-pioneer reopened this Oct 12, 2023
@cyber-pioneer cyber-pioneer changed the title Llama ops [Prim][PIR] Support composite rules of Llama ops Oct 13, 2023
@phlrain phlrain self-assigned this Oct 13, 2023
@phlrain phlrain self-requested a review October 13, 2023 06:51
XiaoguangHu01
XiaoguangHu01 approved these changes Oct 13, 2023
View reviewed changes
Copy link
Contributor

@XiaoguangHu01 XiaoguangHu01 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@cyber-pioneer cyber-pioneer merged commit daa154a into PaddlePaddle:develop Oct 13, 2023
Frida-a pushed a commit to Frida-a/Paddle that referenced this pull request Oct 14, 2023
@cyber-pioneer @Frida-a
[Prim][PIR] Support composite rules of Llama ops (PaddlePaddle#58018)
adf59a0 
* pir prim support decomposite rule of ops

* PIR Prim support ops

* change unnecessary error raise

* fix code

* fix code

* support prim op register

* support max vjp

* fix code

* fixx test case
jiahy0825 pushed a commit to jiahy0825/Paddle that referenced this pull request Oct 16, 2023
@cyber-pioneer @jiahy0825
[Prim][PIR] Support composite rules of Llama ops (PaddlePaddle#58018)
08a437f 
* pir prim support decomposite rule of ops

* PIR Prim support ops

* change unnecessary error raise

* fix code

* fix code

* support prim op register

* support max vjp

* fix code

* fixx test case
danleifeng pushed a commit to danleifeng/Paddle that referenced this pull request Nov 14, 2023
@cyber-pioneer
[Prim][PIR] Support composite rules of Llama ops (PaddlePaddle#58018)
e70e8a5 
* pir prim support decomposite rule of ops

* PIR Prim support ops

* change unnecessary error raise

* fix code

* fix code

* support prim op register

* support max vjp

* fix code

* fixx test case
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@YuanRisheng YuanRisheng YuanRisheng approved these changes

@phlrain phlrain phlrain approved these changes

@XiaoguangHu01 XiaoguangHu01 XiaoguangHu01 approved these changes

@Charles-hit Charles-hit Charles-hit approved these changes

@xiaoguoguo626807 xiaoguoguo626807 xiaoguoguo626807 approved these changes

Assignees

@phlrain phlrain

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
6 participants
@cyber-pioneer @YuanRisheng @phlrain @XiaoguangHu01 @Charles-hit @xiaoguoguo626807