All pull requests

Pull requests

• formers

  stale
  #10732·lugimzzz opened Jun 12, 2025Contributor·

  ·
  4

• [Auto Parallel] Add gpt pp model in dynamic auto

  stale
  #10731·Xing-lil opened Jun 12, 2025Contributor·2 tasks·

  ·
  4

• [Autoparallel] support llama sep

  stale
  #10730·AndSonder opened Jun 12, 2025Contributor·1 of 2 tasks·

  ·
  5

• [x86 CPU] fix cpu GQA issue

  contributorstale
  #10729·yanfeich opened Jun 12, 2025Contributor·2 tasks·

  ·
  5

  

• [DeepGEMM] Print tuning and compilation time stat

  stale
  #10725·chen2016013 opened Jun 11, 2025Contributor·2 tasks·

  ·
  4

• update deep_gemm

  stale
  #10724·chen2016013 opened Jun 10, 2025Contributor·2 tasks·

  ·
  6

• 使用新的aistudio sdk下载模型文件，上传功能不受影响

  contributorstale
  #10723·l1l1l1l opened Jun 10, 2025·

  ·
  7

  

• [Auto-Parallel] Add benchmark for fast_rms_norm in llama13b N4C32 dy_auto

  stale
  #10713·Xing-lil opened Jun 9, 2025Contributor·2 tasks·

  ·
  6

• add register for auto model

  contributorstale
  #10699·zhao-xi opened Jun 4, 2025·1 of 2 tasks·

  ·
  5

  

• Llama auto 13b benchmark with fuse linear

  stale
  #10695·waliwali777 opened Jun 3, 2025Contributor·2 tasks·

  ·
  5

• Tmp llama13b benchmark

  stale
  #10694·liym27 opened Jun 3, 2025Contributor·2 tasks·

  ·
  4

• [feat] Integrate Galvatron (an automatic parallel system integrating …

  contributorstale
  #10680·Time-has-wings opened May 30, 2025·

  ·
  5

  

• Apply merge and split function for mp_reshard

  contributorstale
  #10637·fjjF77 opened May 22, 2025Contributor·

  ·
  5

  

• [Template] add unified template.

  stale
  #10633·wtmlon opened May 22, 2025Collaborator·

  ·
  6

• Remove baichuan13b hidden layers in dynamic auto mode

  stale
  #10624·Function-Samuel opened May 21, 2025Contributor·2 tasks·

  ·
  5

• support_dynamic_server for wintx

  stale
  #10589·chang-wenbin opened May 13, 2025Contributor·2 tasks·

  ·
  5

• [Performance] Boost paddlenlp_ops Compiling

  contributorstale
  #10579·l1cacheDell opened May 12, 2025Contributor·2 tasks·

  ·
  7

  

• [Inference] Add new wint2.75/wint2.5 quant type and support DeepseekV3

  stale
  #10578·lixcli opened May 9, 2025Contributor·

  ·
  13

• 【Inference】support rl

  stale
  #10567·gzy19990617 opened May 8, 2025Contributor·2 tasks·

  ·
  5

• [RL] Refoctor code and add comments

  stale
  #10565·DrownFish19 opened May 7, 2025Collaborator·2 tasks done·

  ·
  7

• [LLM] Modify fuse layout

  stale
  #10555·SylarTiaNII opened May 7, 2025Contributor·

  ·
  4

• Add llama-13b dynamic auto benchmark

  stale
  #10549·Function-Samuel opened May 7, 2025Contributor·2 tasks·

  ·
  4

• Add gpt3 13b dynamic auto benchmark

  stale
  #10548·Function-Samuel opened May 7, 2025Contributor·2 tasks·

  ·
  4

• [LLM] add fuse attention options to LlmMetaConfig

  stale
  #10541·SylarTiaNII opened May 6, 2025Contributor·

  ·
  5

• add append_attn ut

  #10527·bukejiyu opened Apr 29, 2025Contributor·2 tasks·

  ·
  4