All pull requests

Pull requests

• update: llama3

  #556·Lusfie opened Oct 23, 2024·

  ·
  4

• feat: enable ModelLoaderHuggerFace to support loading models in fp16 for inference

  #555·0x404 opened Sep 22, 2024Contributor·

  ·
  1

• Dev chat_glm npu

  #544·xiaohoua opened Aug 21, 2024Contributor·

  ·
  1

• Dev llama npu

  #543·Flowingsun007 opened Aug 19, 2024·

  ·

• support new device

  #540·fpzh2011 opened Jun 28, 2024Contributor·

  ·

• fix path

  #533·zsw256 opened Mar 1, 2024Contributor·

  ·

• add check_meta and sync_data for to_global

  #530·hanwen-sun opened Feb 3, 2024·

  ·

• add gpt2 based llama2

  #529·loxs123 opened Jan 22, 2024Contributor·

  ·
  1

• 固定tensorboardX版本

  #511·L-Xiafeng opened Jun 25, 2023Contributor·

  ·

• Selective checkpointing

  #507·mosout opened May 14, 2023·

  ·

• Dtr exp selective checkpointing

  #506·mosout opened May 13, 2023·

  ·

• Dtr exp

  #503·mosout opened May 7, 2023·

  ·

• Dev alpaca lora

  #499·CPFLAME opened Apr 19, 2023Contributor·

  ·

• mock model training demo【dont merge】

  #497·xiezipeng-ML opened Apr 18, 2023Contributor·

  ·
  1

• fix-loader-fp16

  #490·zhangfanTJU opened Apr 5, 2023Contributor·

  ·
  2

• refine device type

  #488·xiezipeng-ML opened Apr 1, 2023Contributor·

  ·
  1

• 【Dont Merge】cambricon support gpt2 train and inference

  #482·BBuf opened Mar 23, 2023Contributor·

  ·

• Add support for dynamic control flow of auto placement.

  #476·SanftMonster opened Mar 15, 2023·

  ·
  2

• Dev mock pipeline inference

  #474·CPFLAME opened Mar 13, 2023Contributor·

  ·

• test occl

  #471·Panlichen opened Mar 9, 2023·

  ·

• use fmha

  #469·liujuncheng opened Mar 6, 2023Contributor·

  ·

• Fix eod token

  #460·zhangfanTJU opened Feb 19, 2023Contributor·

  ·

• Fix t5 dataset

  #459·zhangfanTJU opened Feb 19, 2023Contributor·

  ·

• fix-mt5-mask-dropout

  #458·zhangfanTJU opened Feb 18, 2023Contributor·

  ·

• try1

  #452·pomegranate9 opened Jan 17, 2023·

  ·