Skip to content

Releases: dkubatin/clearml-darknet-py

v0.2.1

09 Jun 10:58

Choose a tag to compare

Update project documentation. (#12 )

v0.2.0

08 Jun 15:10

Choose a tag to compare

Features

  • Add a parameter to regulate the save weights. (#9)
  • Add the option to save weights every N iterations. (#8)
  • Add the parameter of saving weights from the beginning of N iteration. (#8)

Fixes

  • Fix buffer size closure, add set_subprocess_buffer_size function. (#7)
  • Fix a bug in the save weights. (#2)

Improvements

  • Update version clearml. (#10)

Docs

  • Add autotests.

v0.1.3

06 May 15:11

Choose a tag to compare

Features

  • Add parsing the value of the learning rate parameter with the output of the graph in clearml. (#6)

v0.1.2

06 May 11:40

Choose a tag to compare

Fixes

  • Add support for splitting a dataset from a data hierarchy. (#4)
  • Move additional functions to utils. (#5)

v0.1.1

05 May 14:45

Choose a tag to compare

Features

  • Add hyperparameter parsing of the neural network configuration. (#1)
  • Add network hyperparameter output to cleaml. (#1)
  • Add support for managing network hyperparameters via clearml. (#1)

Fixes

  • Complementing regular expressions to search for -nan values ​​in a string. (#2)
  • Minor fixes to tests. (#3)

Docs

  • Change package usage examples. (#3)
  • Add an example using the detector optimizer. (#3)

v0.1.0

05 May 14:22

Choose a tag to compare

First release of the library!