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

[LLM Inference] Support Qwen2_Moe Inference with MultiGPU #9121

Merged
merged 5 commits into from
Sep 12, 2024

Conversation

CJ77Qi
Copy link
Contributor

@CJ77Qi CJ77Qi commented Sep 11, 2024

PR types

New features

PR changes

Models

Description

支持Qwen/Qwen2-57B-A14B 多卡推理

Copy link

paddle-bot bot commented Sep 11, 2024

Thanks for your contribution!

Copy link

codecov bot commented Sep 11, 2024

Codecov Report

Attention: Patch coverage is 0% with 39 lines in your changes missing coverage. Please review.

Project coverage is 53.32%. Comparing base (51b54d2) to head (4b10081).
Report is 1 commits behind head on develop.

Files with missing lines Patch % Lines
...lp/experimental/transformers/qwen2_moe/modeling.py 0.00% 30 Missing ⚠️
...erimental/transformers/fused_transformer_layers.py 0.00% 9 Missing ⚠️
Additional details and impacted files
@@             Coverage Diff             @@
##           develop    #9121      +/-   ##
===========================================
- Coverage    53.33%   53.32%   -0.02%     
===========================================
  Files          652      652              
  Lines       105404   105436      +32     
===========================================
+ Hits         56222    56225       +3     
- Misses       49182    49211      +29     

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

@@ -832,26 +866,34 @@ def get_tensor_parallel_split_mappings(num_layers):
# Row Linear
"embed_tokens.weight": partial(fn, is_column=False),
"layers.0.self_attn.o_proj.weight": partial(fn, is_column=False),
"layers.0.mlp.down_proj.weight": partial(fn, is_column=False),
# "layers.0.mlp.down_proj.weight": partial(fn, is_column=False),
Copy link
Collaborator

Choose a reason for hiding this comment

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

这条注释给删掉吧

@qingqing01 qingqing01 merged commit d3302c5 into PaddlePaddle:develop Sep 12, 2024
5 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