Commits

Breadcrumbs

History for

ondevelop

User selector

All users

All time

Commit History

Commits on Jul 1, 2024

• bug fix (#8687)

  wtmlon

  authored
  2723138

  Copy full SHA for 2723138

Commits on Jun 28, 2024

• Sft flash mask (#8664)

  Show description for c86e447
  

  wtmlon

  and

  gongel

  authored
  c86e447

  Copy full SHA for c86e447

Commits on Jun 27, 2024

• Update export model (#8671)

  Show description for cd5ca82
  DesmonDay

  authored
  cd5ca82

  Copy full SHA for cd5ca82

• adjust llm readme (#8672)

  Show description for 513e21d
  lugimzzz

  authored
  513e21d

  Copy full SHA for 513e21d

• support llama avx model inference (#8634)

  bukejiyu

  authored
  69e04c0

  Copy full SHA for 69e04c0

• inference support llama3(wint8|4/a8w8) (#8630)

  Show description for faabf87
  yuanlehome

  authored
  faabf87

  Copy full SHA for faabf87

Commits on Jun 26, 2024

• [DOC] Fix dead link (#8662)

  Show description for 4e971f8
  DrownFish19

  authored
  4e971f8

  Copy full SHA for 4e971f8

• [LLM][NPU] fix on readme (#8659)

  SylarTiaNII

  authored
  8edb84e

  Copy full SHA for 8edb84e

Commits on Jun 25, 2024

• align llama auto_parallel dataloader with manual_parallel (#8639)

  zhiqiu

  authored
  11a7fd3

  Copy full SHA for 11a7fd3

Commits on Jun 22, 2024

• [NPU][LLM] add README & reformat llama scripts (#8642)

  Show description for 25d2140
  SylarTiaNII

  authored
  25d2140

  Copy full SHA for 25d2140

Commits on Jun 21, 2024

• [LLM] change llm content (#8627)

  Show description for 498f709
  

  lugimzzz

  and

  zjjlivein

  authored
  498f709

  Copy full SHA for 498f709

Commits on Jun 20, 2024

• [LLM] Support prefix tuning and lora for qwen2 (#8601)

  Show description for da8b9ac
  DrownFish19

  authored
  da8b9ac

  Copy full SHA for da8b9ac

• [PaddleNLP 3.0] update dataset preprocess (#8629)

  Show description for 6bfca91
  DrownFish19

  authored
  6bfca91

  Copy full SHA for 6bfca91

Commits on Jun 19, 2024

• [PaddleNLP 3.0] Refactor RLHF (#8617)

  Show description for 4e16650
  gongel

  authored
  4e16650

  Copy full SHA for 4e16650

Commits on Jun 18, 2024

• Finetune support continue_training (#8615)

  tianhaodongbd

  authored
  cbd0ab3

  Copy full SHA for cbd0ab3

Commits on Jun 14, 2024

• [Feature] Optimize config saving. (#8490)

  Show description for 0844a5b
  ZHUI

  authored
  0844a5b

  Copy full SHA for 0844a5b

Commits on Jun 13, 2024

• [Bug Fix] fix sharding stage1 allgather overlap bug, which needs to forbiden pin memory (#8594)

  Show description for 439f8f3
  iosmers

  authored
  439f8f3

  Copy full SHA for 439f8f3

• [XPU] Open ir pass (#8598)

  zhink

  authored
  2dc4d7b

  Copy full SHA for 2dc4d7b

• Fix dataset with empty char. (#8469)

  Show description for 8e24341
  ZHUI

  authored
  8e24341

  Copy full SHA for 8e24341

• [XPU] xpu devices support llama-7b basic mode inference (turn on BlockAttention) (#8588)

  Show description for 3d777c1
  zhink

  authored
  3d777c1

  Copy full SHA for 3d777c1

Commits on Jun 11, 2024

• [LLM] support Qwen2 (#8338)

  Show description for 4609d07
  DrownFish19

  authored
  4609d07

  Copy full SHA for 4609d07

• add llama & qwen dpo (#8474)

  Show description for 909be01
  lugimzzz

  authored
  909be01

  Copy full SHA for 909be01

Commits on Jun 6, 2024

• Update llama_npu_opt_lora.sh (#8562)

  Galaxy1458

  authored
  4e3f60d

  Copy full SHA for 4e3f60d

Commits on Jun 5, 2024

• Add RingFlashAttention for context parallel (#8383)

  Show description for 4cb8b6d
  zhangyuqin1998

  authored
  4cb8b6d

  Copy full SHA for 4cb8b6d

Commits on Jun 3, 2024

• Update README.md (#8524)

  wawltor

  authored
  fd2a39e

  Copy full SHA for fd2a39e

Commits on May 31, 2024

• Add Sharding V1 broadcast and allgather overlap optimize (#8499)

  Show description for e71540b
  

  3 peopleauthored
  e71540b

  Copy full SHA for e71540b

• Update merge_lora_params.py (#8514)

  Show description for 60c6615
  Galaxy1458

  authored
  60c6615

  Copy full SHA for 60c6615

Commits on May 28, 2024

• [AutoParallel] Support qwen for auto_parallel (#8312)

  Show description for 85ba573
  

  GhostScreaming

  and

  pangengzheng

  authored
  85ba573

  Copy full SHA for 85ba573

Commits on May 27, 2024

• Fix rng_state in llm models (#8396)

  zhangyuqin1998

  authored
  d4edd19

  Copy full SHA for d4edd19

Commits on May 20, 2024

• [NPU] support npu llama2-13B export & inference (#8442)

  Show description for 87e4c4f
  ronny1996

  authored
  87e4c4f

  Copy full SHA for 87e4c4f

Commits on May 17, 2024

• [benchmark]Add llama2 auto (#8424)

  Show description for 5d4ce56
  Liujie0926

  authored
  5d4ce56

  Copy full SHA for 5d4ce56

• update (#8448)

  Show description for 7f13396
  Galaxy1458

  authored
  7f13396

  Copy full SHA for 7f13396

Commits on May 16, 2024

• Support fused_attention_qkv for auto_parallel llama (#8432)

  Show description for ebe397e
  zhangbo9674

  authored
  ebe397e

  Copy full SHA for ebe397e

Commits on May 15, 2024

• fix npu sft ckpt load bug and no FA bug (#8438)

  Show description for 5170664
  

  NINGBENZHE

  and

  NINGBENZHE

  authored
  5170664

  Copy full SHA for 5170664

• Update llama_npu_opt_lora.sh (#8439)

  Galaxy1458

  authored
  11aba32

  Copy full SHA for 11aba32