Skip to content

Releases

v3.0.2

Choose a tag to compare

@bararchy bararchy released this 31 Dec 09:30
  • Fixed regression in CPU path
  • Fixed and Expanded autograd
  • Fixed adam issues

v3.0.1

Choose a tag to compare

@bararchy bararchy released this 08 Jul 17:56
13fc280

What's Changed

Full Changelog: v3.0.0...v3.0.1

v3.0.0

Choose a tag to compare

@bararchy bararchy released this 07 Jul 13:37
45ef8d4
  • CUDA Usage
  • Matrix approach
  • LLM possibility for training
  • See new README for full info

v2.5.0

Choose a tag to compare

@bararchy bararchy released this 23 Jun 19:03
6d34991

What's Changed

Full Changelog: v2.4.0...v2.5.0

v2.4.0

Choose a tag to compare

@bararchy bararchy released this 10 Jan 17:58

Make it work with crystal 1.6.2 🎉
syntax fixes, cleanup and CI

v2.3.0

Choose a tag to compare

@bararchy bararchy released this 19 Jun 11:09
319a77f
  • Added improved discovery of exploding gradients
  • Added better error report
  • Added autosave during training
  • General refactoring

v2.2.1

Choose a tag to compare

@bararchy bararchy released this 17 Jun 11:48
0fc1b5f
  • Support for Crystal 0.25.0

v2.2.0

Choose a tag to compare

@bararchy bararchy released this 14 Jun 14:30
63fa13e
  • ES Optimizer
  • cross-entropy
  • bug fixes

v2.1.0

Choose a tag to compare

@bararchy bararchy released this 14 May 14:41
2614f91
  • Spec fixes
  • CNN Additions + fixes
  • More overall customizability

v2.0.0

Choose a tag to compare

@bararchy bararchy released this 21 Jan 11:07
ddab78d
  • CNN
  • Bug fixes