Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 16, 2024

• tiny fix&support llava-1.6 models during sft training.

  wangclnlp

  committed
  b6a1f0c

  Copy full SHA for b6a1f0c

Commits on Sep 28, 2024

• tiny fix&support for training the LLaMA-3.2-Vision.

  wangclnlp

  committed
  f880e96

  Copy full SHA for f880e96

Commits on Aug 30, 2024

• support for training a reward model with the LLaVA-NeXT model

  wangclnlp

  committed
  455b86f

  Copy full SHA for 455b86f

Commits on Aug 23, 2024

• fix some bugs, and support for training LLaVA-NeXT.

  wangclnlp

  committed
  f704865

  Copy full SHA for f704865

Commits on Jul 21, 2024

• fix some bugs.

  wangclnlp

  committed
  427f289

  Copy full SHA for 427f289

Commits on Jul 18, 2024

• Cleaning codes for processing multiple images.

  wangclnlp

  committed
  63d424b

  Copy full SHA for 63d424b

Commits on Jul 10, 2024

• Support loading of a LLaVA model in all training stages & fix some bugs.

  wangclnlp

  committed
  80360c6

  Copy full SHA for 80360c6

Commits on Jul 7, 2024

• Support loading of a LLaVA model during the SFT training phase.

  wangclnlp

  committed
  674bc36

  Copy full SHA for 674bc36

Commits on Jul 5, 2024

• tiny fix.

  wangclnlp

  committed
  274a417

  Copy full SHA for 274a417

• add pretraining stage.

  wangclnlp

  committed
  058a346

  Copy full SHA for 058a346

Commits on Jul 2, 2024

• DPO training with multiple candidate outputs is supported.

  wangclnlp

  committed
  67bf6dd

  Copy full SHA for 67bf6dd

• Some improvemants for this system.

  wangclnlp

  committed
  068fcf9

  Copy full SHA for 068fcf9

Commits on Jul 1, 2024

• first commit

  wangclnlp

  committed
  0a2678f

  Copy full SHA for 0a2678f

Loading