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.

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

rename files and add readme for llama auto_parallel #7944

Merged
merged 3 commits into from
Feb 1, 2024

Conversation

zhiqiu
Copy link
Collaborator

@zhiqiu zhiqiu commented Jan 31, 2024

PR types

Others

PR changes

Docs

Description

rename files and add readme for llama auto_parallel

Copy link

codecov bot commented Jan 31, 2024

Codecov Report

Attention: 150 lines in your changes are missing coverage. Please review.

Comparison is base (3a704ea) 56.58% compared to head (d89e933) 56.58%.

Files Patch % Lines
...ddlenlp/transformers/llama/modeling_auto_static.py 4.76% 100 Missing ⚠️
paddlenlp/transformers/llama/modeling_auto.py 9.09% 50 Missing ⚠️
Additional details and impacted files
@@           Coverage Diff            @@
##           develop    #7944   +/-   ##
========================================
  Coverage    56.58%   56.58%           
========================================
  Files          589      589           
  Lines        89900    89900           
========================================
+ Hits         50870    50872    +2     
+ Misses       39030    39028    -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
)
# loss = None
Copy link
Collaborator

Choose a reason for hiding this comment

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

注释的直接删除?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

后面提PR完善代码

layer.down_proj.weight.scale_(factor)
if isinstance(layer, LlamaAttentionAuto):
factor = 1 / math.sqrt(2 * self.config.num_hidden_layers)
layer.o_proj.weight.scale_(factor)
Copy link
Collaborator

Choose a reason for hiding this comment

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

删除?

# masked_lm_loss = self.loss_func(prediction_scores.astype("float32"), masked_lm_labels.unsqueeze(2))
# # skip ignore_index which loss == 0
# masked_lm_loss = masked_lm_loss[masked_lm_loss > 0].astype("float32")
# loss = paddle.mean(masked_lm_loss)
Copy link
Collaborator

Choose a reason for hiding this comment

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

删除?

Copy link
Collaborator

@ZHUI ZHUI left a comment

Choose a reason for hiding this comment

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

LGTM

@wawltor wawltor merged commit 8d6e813 into PaddlePaddle:develop Feb 1, 2024
7 of 10 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants