Recent Releases of trainer

trainer - v0.0.36

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.35...v0.0.36

- Python
Published by erogol over 2 years ago

trainer - v0.0.35

What's Changed

  • Revert "Revert "fix: make --continue_path work again (#131)"" by @eginhard in https://github.com/coqui-ai/Trainer/pull/135
  • Fixup savebestmodel by @erogol in https://github.com/coqui-ai/Trainer/pull/137

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.34...v0.0.35

- Python
Published by erogol over 2 years ago

trainer - v0.0.34

What's Changed

  • Revert "fix: make --continue_path work again" by @erogol in https://github.com/coqui-ai/Trainer/pull/133

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.33...v0.0.34

- Python
Published by erogol over 2 years ago

trainer - v0.0.33

What's Changed

  • fix: make --continue_path work again by @eginhard in https://github.com/coqui-ai/Trainer/pull/131

New Contributors

  • @eginhard made their first contribution in https://github.com/coqui-ai/Trainer/pull/131

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.32...v0.0.33

- Python
Published by erogol over 2 years ago

trainer - v0.0.32

What's Changed

  • bugfix: distribute rank initialization by @Squire-tomsk in https://github.com/coqui-ai/Trainer/pull/126
  • Multiples bug fixes and add ontrainepoch_start callback by @Edresson in https://github.com/coqui-ai/Trainer/pull/129

New Contributors

  • @Squire-tomsk made their first contribution in https://github.com/coqui-ai/Trainer/pull/126

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.31...v0.0.32

- Python
Published by erogol over 2 years ago

trainer - v0.0.31

What's Changed

  • Handle no eval loss by @erogol in https://github.com/coqui-ai/Trainer/pull/121
  • Bump up to v0.0.31 by @erogol in https://github.com/coqui-ai/Trainer/pull/122

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.30...v0.0.31

- Python
Published by erogol almost 3 years ago

trainer - v0.0.30

What's Changed

  • Update trainer.py by @ppisljar in https://github.com/coqui-ai/Trainer/pull/118

New Contributors

  • @ppisljar made their first contribution in https://github.com/coqui-ai/Trainer/pull/118

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.29...v0.0.30

- Python
Published by erogol almost 3 years ago

trainer - v0.0.29

What's Changed

  • Fix accelerate init by @erogol in https://github.com/coqui-ai/Trainer/pull/116

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.28...v0.0.29

- Python
Published by erogol about 3 years ago

trainer - v0.0.28

What's Changed

  • Accelerate by @erogol in https://github.com/coqui-ai/Trainer/pull/114

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.27...v0.0.28

- Python
Published by erogol about 3 years ago

trainer - v0.0.27

What's Changed

  • Update to p3.11 by @erogol in https://github.com/coqui-ai/Trainer/pull/111

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.26...v0.0.27

- Python
Published by erogol about 3 years ago

trainer - v0.0.26

What's Changed

  • bugfix: fix getlastcheckpoint on windows by @Zutatensuppe in https://github.com/coqui-ai/Trainer/pull/107
  • Enable bfloat16 training by @erogol in https://github.com/coqui-ai/Trainer/pull/108

New Contributors

  • @Zutatensuppe made their first contribution in https://github.com/coqui-ai/Trainer/pull/107

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.25...v0.0.26

- Python
Published by erogol about 3 years ago

trainer - v0.0.25

What's Changed

  • Bug fix on cpu autocast by @Edresson in https://github.com/coqui-ai/Trainer/pull/103
  • free outputs in training loop as soon as possible as it makes a big difference in constrained-memory scenarios by @shenberg in https://github.com/coqui-ai/Trainer/pull/104

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.24...v0.0.25

- Python
Published by erogol over 3 years ago

trainer - v0.0.24

What's Changed

  • minimal fix to wandb_logger.py necessary to get Overflow running using wandb by @shenberg in https://github.com/coqui-ai/Trainer/pull/98

New Contributors

  • @shenberg made their first contribution in https://github.com/coqui-ai/Trainer/pull/98

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.23...v0.0.24

- Python
Published by erogol over 3 years ago

trainer - v0.0.23

What's Changed

  • Update logging by @erogol in https://github.com/coqui-ai/Trainer/pull/100
  • Add option to pass custom loaders by @loganhart420 in https://github.com/coqui-ai/Trainer/pull/92
  • Save on exit by @erogol in https://github.com/coqui-ai/Trainer/pull/102

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.22...v0.0.23

- Python
Published by erogol over 3 years ago

trainer - v0.0.22

What's Changed

  • Optimizer dict by @erogol in https://github.com/coqui-ai/Trainer/pull/94

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.21...v0.0.22

- Python
Published by erogol over 3 years ago

trainer - v0.0.21

What's Changed

  • Custom optimize for handling complex trainings by @erogol in https://github.com/coqui-ai/Trainer/pull/89

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.20...v0.0.21

- Python
Published by erogol over 3 years ago

trainer - v0.0.20

What's Changed

  • Fix loss detach by @erogol in https://github.com/coqui-ai/Trainer/pull/87

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.19...v0.0.20

- Python
Published by erogol over 3 years ago

trainer - v0.0.19

What's Changed

  • Skip and warn when the optimizer does not match by @erogol in https://github.com/coqui-ai/Trainer/pull/78
  • Start with eval by @erogol in https://github.com/coqui-ai/Trainer/pull/86

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.18...v0.0.19

- Python
Published by erogol over 3 years ago

trainer - v0.0.18

What's Changed

  • Add start with eval option by @erogol in https://github.com/coqui-ai/Trainer/pull/84

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.17...v0.0.18

- Python
Published by erogol over 3 years ago

trainer - v0.0.17

What's Changed

  • Add the Largest batch size finder by @loganhart420 in https://github.com/coqui-ai/Trainer/pull/79

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.16...v0.0.17

- Python
Published by erogol over 3 years ago

trainer - v0.0.16

What's Changed

  • Add checkpoint loading warning by @WeberJulian in https://github.com/coqui-ai/Trainer/pull/73
  • Custom callback by @erogol in https://github.com/coqui-ai/Trainer/pull/74

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.15...v0.0.16

- Python
Published by erogol almost 4 years ago

trainer - v0.0.15

What's Changed

  • Restore the best loss just for --continue_path by @Edresson in https://github.com/coqui-ai/Trainer/pull/67
  • Print more info about training environment by @erogol in https://github.com/coqui-ai/Trainer/pull/66

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.14...v0.0.15

- Python
Published by erogol almost 4 years ago

trainer - v0.0.14

What's Changed

  • Fix Rank Zero Only Behavior by @iprovalo in https://github.com/coqui-ai/Trainer/pull/58

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.13...v0.0.14

- Python
Published by erogol almost 4 years ago

trainer - v0.0.13

What's Changed

  • Num gpus fix by @iprovalo in https://github.com/coqui-ai/Trainer/pull/47
  • Profiling by @erogol in https://github.com/coqui-ai/Trainer/pull/49
  • fix for loss not present in avg val by @manmay-nakhashi in https://github.com/coqui-ai/Trainer/pull/50
  • Log training script by @erogol in https://github.com/coqui-ai/Trainer/pull/57

New Contributors

  • @iprovalo made their first contribution in https://github.com/coqui-ai/Trainer/pull/47
  • @manmay-nakhashi made their first contribution in https://github.com/coqui-ai/Trainer/pull/50

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.12...v0.0.13

- Python
Published by erogol about 4 years ago

trainer - v0.0.12

What's Changed

  • Simple fit function for torch profiling by @erogol in https://github.com/coqui-ai/Trainer/pull/44

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.11...v0.0.12

- Python
Published by erogol about 4 years ago

trainer - v0.0.11

What's Changed

  • Update for p3.10 by @erogol in https://github.com/coqui-ai/Trainer/pull/38

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.10...v0.0.11

- Python
Published by erogol about 4 years ago

trainer - v0.0.10

Just to fix the mismatch to the PyPI release version

- Python
Published by erogol about 4 years ago

trainer - v0.0.9

What's Changed

  • Windows Multiprocessing Fix by @bitnom in https://github.com/coqui-ai/Trainer/pull/32
  • Implement setup_training_environment by @erogol in https://github.com/coqui-ai/Trainer/pull/36

New Contributors

  • @bitnom made their first contribution in https://github.com/coqui-ai/Trainer/pull/32

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.8...v0.0.9

- Python
Published by erogol about 4 years ago

trainer - v0.0.8

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.7...v0.0.8

  • Set default cudnn_deterministic False

- Python
Published by erogol over 4 years ago

trainer - v0.0.7

What's Changed

  • Add small run arg by @WeberJulian in https://github.com/coqui-ai/Trainer/pull/24
  • Fix logger use by @erogol in https://github.com/coqui-ai/Trainer/pull/30

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.6...v0.0.7

- Python
Published by erogol over 4 years ago

trainer - v0.0.6

What's Changed

  • Fix continue training and new_fields writing by @WeberJulian in https://github.com/coqui-ai/Trainer/pull/18
  • Rename PyPI package: coqui_trainer --> trainer by @erogol in https://github.com/coqui-ai/Trainer/pull/27
  • Fix DDP training by @erogol in https://github.com/coqui-ai/Trainer/pull/28

New Contributors

  • @WeberJulian made their first contribution in https://github.com/coqui-ai/Trainer/pull/18

Full Changelog: https://github.com/coqui-ai/Trainer/compare/v0.0.5...v0.0.6

- Python
Published by erogol over 4 years ago

trainer - v0.0.5

What's Changed

  • Add Distributed Sampler Wrapper to enable the use of samplers with DDP by @Edresson in https://github.com/coqui-ai/Trainer/pull/3
  • Fix missing requirements by @erogol in https://github.com/coqui-ai/Trainer/pull/8
  • Issue_templates by @erogol in https://github.com/coqui-ai/Trainer/pull/9
  • Add basic CI by @erogol in https://github.com/coqui-ai/Trainer/pull/10
  • Fix saving checkpoints by @erogol in https://github.com/coqui-ai/Trainer/pull/13

New Contributors

  • @Edresson made their first contribution in https://github.com/coqui-ai/Trainer/pull/3

Full Changelog: https://github.com/coqui-ai/Trainer/commits/v0.0.5

- Python
Published by erogol over 4 years ago