Skip to content

Releases: neptune-ai/neptune-pytorch

3.0.0

13 Oct 12:42
2c4d7b5

Choose a tag to compare

What's Changed

Important

Neptune 3.x only: This version only supports Neptune 3.x and requires a Neptune 3.x API token and the neptune-scale API client.
Neptune 2.x users should continue using neptune-pytorch<3.0.0.

Check the CHANGELOG for details.

New Contributors

Full Changelog: 2.0.0...3.0.0

2.0.0

21 Aug 12:32
49f9749

Choose a tag to compare

Changes

  • Rename save_model to log_model and save_checkpoint to log_checkpoint. (#9)
  • Prefix private methods with underscore. (#12)
  • Add docstrings for log_model and log_checkpoint. (#11)

1.1.0

12 Aug 16:13
e129e10

Choose a tag to compare

Changes

  • Rename save_model to log_model and save_checkpoint to log_checkpoint. (#9)

1.0.1

28 Jul 12:03
52628a8

Choose a tag to compare

Fixes

  • Make torchviz optional dependency. (#8)

1.0.0

28 Jun 08:59
bcd2126

Choose a tag to compare

neptune-pytorch 1.0.0

Fixes

  • Change where checkpoints are logged. Previously they we logged under base_namespace/model but now they will be logged under base_namespace/model/checkpoints (#5)
  • Add warning if dot is not installed instead of hard error. Also, improve clean-up of visualization files (#6)

Features

  • Create NeptuneLogger for logging metadata (#1)

neptune-pytorch 0.2.0

05 May 19:07
e51df64

Choose a tag to compare

neptune-pytorch 0.2.0

Fixes

  • Change where checkpoints are logged. Previously they we logged under base_namespace/model but now they will be logged under base_namespace/model/checkpoints (#5)
  • Add warning if dot is not installed instead of hard error. Also, improve clean-up of visualization files (#6)

neptune-pytorch 0.1.0

02 May 09:45
bb1b306

Choose a tag to compare

Create NeptuneLogger for logging metadata when using PyTorch for training.