Skip to content

Conversation

@cjw-d
Copy link
Contributor

@cjw-d cjw-d commented Nov 26, 2025

PR types

Others

PR changes

Others

Description

  • standardize_rope_params和rope_config_validation覆盖新增模型
  • 修改有误的注释
  • 将模型组网内的view修改为reshape

@paddle-bot
Copy link

paddle-bot bot commented Nov 26, 2025

Thanks for your contribution!

@cjw-d cjw-d changed the title cover newly added models cover newly added models and improve tensor reshaping Nov 27, 2025
@codecov-commenter
Copy link

Codecov Report

❌ Patch coverage is 72.28916% with 23 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@d2014a2). Learn more about missing BASE report.

Files with missing lines Patch % Lines
paddleformers/transformers/phi3/configuration.py 60.71% 11 Missing ⚠️
...eformers/transformers/gemma3_text/configuration.py 50.00% 5 Missing ⚠️
paddleformers/transformers/llama/modeling.py 33.33% 4 Missing ⚠️
paddleformers/transformers/qwen2_5_vl/modeling.py 78.57% 3 Missing ⚠️

❌ Your patch status has failed because the patch coverage (72.28%) is below the target coverage (80.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files
@@            Coverage Diff             @@
##             develop    #3026   +/-   ##
==========================================
  Coverage           ?   33.55%           
==========================================
  Files              ?      428           
  Lines              ?    69474           
  Branches           ?        0           
==========================================
  Hits               ?    23315           
  Misses             ?    46159           
  Partials           ?        0           

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

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

)
distributed_input_to_alltoall_out = (
distributed_input_to_alltoall_out.view([num_local_experts, -1, capacity])
distributed_input_to_alltoall_out.reshape([num_local_experts, -1, capacity])
Copy link
Collaborator

Choose a reason for hiding this comment

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

这是把所有模型里的view都换成reshape吗?没有问题的模型,不需要全部替换为reshape

Copy link
Contributor Author

Choose a reason for hiding this comment

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

好的

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants