Sage Attention For deepseek prefilling. WIP

[l1cacheDell]

PR brief

Sage Attention sm90 cuda kernel specialized for Deepseek mixed head_dim=192 inference.

Now make a demo inference code.

Benchmarking...

[paddle-bot]

Thanks for your contribution!

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 12 lines in your changes missing coverage. Please review.

Project coverage is 51.07%. Comparing base (45b5012) to head (a6ccb70).
Report is 4 commits behind head on develop.

Files with missing lines	Patch %	Lines
...erimental/transformers/fused_transformer_layers.py	0.00%	12 Missing ⚠️

❌ Your patch check has failed because the patch coverage (0.00%) is below the target coverage (80.00%). You can increase the patch coverage or adjust the target coverage.
❌ Your project check has failed because the head coverage (51.07%) is below the target coverage (58.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #9958      +/-   ##
===========================================
- Coverage    51.08%   51.07%   -0.01%     
===========================================
  Files          745      745              
  Lines       119261   119285      +24     
===========================================
+ Hits         60926    60927       +1     
- Misses       58335    58358      +23     

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