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

Xpu 2.10 test=develop #8199

Merged
merged 34 commits into from
Jan 17, 2022
Merged

Xpu 2.10 test=develop #8199

merged 34 commits into from
Jan 17, 2022

Conversation

newway
Copy link
Collaborator

@newway newway commented Jan 5, 2022

cherry-pick develop XPU patches to 2.10 branch, from commit [bac2ca6, d893d7f], test=develop

shanliang1992 and others added 30 commits December 24, 2021 16:41
@shanliang1992 @newway
[XPU] change match_matrix_tensor op from old version to refector veri…
c55f051 
…son (PaddlePaddle#7012)
@shanliang1992 @newway
[XPU] change sequence concat op from old version to refector verison (P…
dac25b4 
…addlePaddle#6847)
@shanliang1992 @newway
[XPU] change sequence_reverse api from old version to refector version (
8b98474 
PaddlePaddle#6798)
@wanglei91 @newway
[XPU] fix some bugs for transformer (PaddlePaddle#7014)
17ed939
@newway
[XPU] Mul quant (PaddlePaddle#6850)
addd8a4
@newway
[XPU] bugfix on fc max (PaddlePaddle#7152)
127b860
@zhupengyang @newway
expand_v2 supports dynamic shape (PaddlePaddle#7116)
6136831
@shanliang1992 @newway
[XPU] change search_noaligned_mat_mul op to fc_batched_vsl op (Paddle…
d7efe3e 
…Paddle#7081)
@newway
[xpu] support qkv-fused weight reuse in scs_tran_match (PaddlePaddle#…
3da529d 
…7293)
@newway
[XPU] Scs trans match (PaddlePaddle#7307)
cdffd68
@zhupengyang @newway
[XPU] add lod_array_length; argmax support int32 (PaddlePaddle#7314)
4dc1c3d 
test=develop
@shanliang1992 @newway
[XPU] change fc_int16 op to fc_fusion (PaddlePaddle#7029)
901c7a8
@mayang002 @newway
[XPU] super big ernie support (PaddlePaddle#7184)
a676214
@mayang002 @newway
[XPU] free default workspace of xpu_ctx before setting up a new gm wo…
3964ce1 
…rkspace (PaddlePaddle#7422)
@Gradie @newway
[XPU] use get_max_ptr function (PaddlePaddle#7482)
b6dd5c2 
Co-authored-by: liuli31 <[email protected]>
@newway
[xpu] Fc int31 (PaddlePaddle#7514)
ef4f397 
* [xpu] fix continuous encoder fuse and fc max size

* [xpu] refactor fc int31 for KL2
@Gradie @newway
use get_max_ptr function (PaddlePaddle#7529)
7fb282e
@Gradie @newway
add activation xpu gelu (PaddlePaddle#7527)
e7f907b
@newway
[xpu] more check with multi_encoder pass (PaddlePaddle#7593)
2b8c321
@shanliang1992 @newway
[XPU]use get_max_ptr_size in search attention op (PaddlePaddle#7528)
7fc88b4
@shanliang1992 @newway
[XPU]use get_max_ptr_size in bigru op (PaddlePaddle#7498)
9cd716b
@shanliang1992 @newway
[XPU] change sequence_topk_avg_pooling op from old version to refecto…
39c4d15 
…r verison (PaddlePaddle#7411)
@Gradie @newway
update xpu api sequence_unpad (PaddlePaddle#7640)
9b7f8cf
@Gradie @newway
update xpu api l2_norm (PaddlePaddle#7724)
5f12281
@zhupengyang @newway
[XPU] __xpu__resnet_fuse_pass should not match ResNeXt50 (PaddlePaddl…
2e2c4a7 
…e#7824)
@newway
[xpu] support encoder mul shape without equal length and more check (P…
cb0e511 
…addlePaddle#7753)
@shanliang1992 @newway
[XPU] use new search_varconv (PaddlePaddle#7865)
88402b0
@shanliang1992 @newway
[XPU] use new sequence_topk_avg_pooling (PaddlePaddle#7834)
70c4e0d
@liyupeng @newway
[XPU] fix xpu memory leak bug for arm arch (PaddlePaddle#8010)
52466b6
@shanliang1992 @newway
[XPU] new new op in mmdnn (PaddlePaddle#7998)
3793513
@paddle-bot-old
Copy link

paddle-bot-old bot commented Jan 5, 2022

Thanks for your contribution!

zhupengyang
zhupengyang approved these changes Jan 17, 2022
View reviewed changes
Copy link
Collaborator

@zhupengyang zhupengyang left a comment

Choose a reason for hiding this comment

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

LGTM

@zhupengyang zhupengyang merged commit bb2edd6 into PaddlePaddle:release/v2.10 Jan 17, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@zhupengyang zhupengyang zhupengyang approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
8 participants
@newway @zhupengyang @shanliang1992 @wanglei91 @mayang002 @Gradie @liyupeng @fc657547899