[Auto Parallel] Support dynamic semi-auto training in Llama2 model #7851
Conversation
This file contains 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
|
Thanks for your contribution! |
haohongxiang
force-pushed
the
dygraph_semi_auto_llama2
branch
from
e16875e
to
de9bb93
Compare
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## develop #7851 +/- ##
===========================================
- Coverage 56.96% 56.67% -0.29%
===========================================
Files 587 588 +1
Lines 88647 89243 +596
===========================================
+ Hits 50494 50580 +86
- Misses 38153 38663 +510 ☔ View full report in Codecov by Sentry. |
haohongxiang
force-pushed
the
dygraph_semi_auto_llama2
branch
15 times, most recently
from
cf7d444
to
b9acdf0
Compare
haohongxiang
force-pushed
the
dygraph_semi_auto_llama2
branch
from
b9acdf0
to
327d788
Compare
haohongxiang
changed the title
wawltor
reviewed
Jan 18, 2024
|
|
||
| auto variance_shape = x_shape; | ||
| variance_shape.pop_back(); | ||
| auto invvar = paddle::empty(variance_shape, paddle::DataType::FLOAT32, place); |
There was a problem hiding this comment.
fused_ln的改动是因为原本variance的infer shape有问题，只是动手不会报错；动半加上切分推导规则，就会挂，所以需要修复。同样的，框架里layer norm算子也做了修复，详见PR-58776
ZHUI
reviewed
Jan 18, 2024
| return outputs | ||
|
|
||
|
|
||
| class LlamaDecoderLayerAuto(nn.Layer): |
There was a problem hiding this comment.
这些模型名字是不是跟 modeling_auto 是冲突的?
There was a problem hiding this comment.
这些模型名字仅用于当前文件内自用，没有加到__all__列表里，不会对用户或训练侧暴露；另外这里只是中间态，后期等动静半执行代码接入后，仅保留modeling_3D_auto.py，原本纯静半的modeling_auto.py会删除
| @@ -14,6 +14,7 @@ | |||
|
|
|||
| from .configuration import * | |||
| from .modeling import * | |||
| from .modeling_3D_auto import * | |||
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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 types
Bug fixes
PR changes
Others
Description
[Auto Parallel] Support dynamic semi-auto training in Llama2 model