v0.6.0

What's Changed

• Add multimodal by @warofgam in #133
• miltimodal-demo by @Dzagcoffee in #140
• Speedup collect_lists and add test by @ivan-chai in #141
• multimodal_demo_replace by @Dzagcoffee in #144
• Demo supervised transformer and sop, nsp embeddings by @biQvOid in #146
• Huggingface datasets by @dllllb in #149 and #150
• transaction-emb.ipynb creating by @IvanSergeevPhysics in #145
• Mlm demo restore by @ivkireev86 in #147
• Update README.md by @Dzagcoffee in #154
• Add multi-gpu by @FuTSy13 in #139
• Modern torch by @ivan-chai in #156

New Contributors

• @warofgam made their first contribution in #133
• @Dzagcoffee made their first contribution in #140
• @ivan-chai made their first contribution in #141
• @biQvOid made their first contribution in #146
• @IvanSergeevPhysics made their first contribution in #145

Full Changelog: v0.5.3...v0.6.0

v0.5.3

What's Changed

• fix multilabel demo by @ArtyomVorobev in #104
• softmax loss update by @ivkireev86 in #102
• pl_fit_target.py fix by @FuTSy13 in #105
• Fix GRU multilayer hidden size init by @morlov in #109
• Dependency update by @dllllb in #111
• Add managing with external data, fine-tuning for external data embedders by @WhatCanWeDo in #110
• fix_trx_encoder_base by @ivkireev86 in #112
• DuckDB dataset by @dllllb in #113
• Fix boolean target by @Dzhambo in #115
• Add time splitters and negative sampling by @Dzhambo in #118
• Add grouped metrics for tensorboard by @Dzhambo in #120
• Update padded_batch.py by @tsebaka in #121
• trx encoder category max size fix by @Ilykuleshov in #125
• add test for trx_encoder_base category max size by @ivkireev86 in #129
• hydra notebook example by @tsebaka in #126
• Text features demo by @ivkireev86 in #116
• How to contribute by @ivkireev86 in #134
• Update trx_encoder_base.py get_custom_embeddings rollback by @ivkireev86 in #136
• coles-emb.ipynb - torch downgrage by @ivkireev86 in #135

New Contributors

• @morlov made their first contribution in #109
• @WhatCanWeDo made their first contribution in #110
• @Dzhambo made their first contribution in #115
• @tsebaka made their first contribution in #121
• @Ilykuleshov made their first contribution in #125

Full Changelog: v0.5.2...v0.5.3

v0.5.2

What's Changed

• Augmentation datasets by @ArtyomVorobev in #94
• gpt_module and custom transformer by @mazitovs in #95
• pyarrow requirement hotfix by @FuTSy13 in #99
• python37 github test by @ivkireev86 in #98
• Multilabel demo and mlm hotfix by @ArtyomVorobev in #100
• Hotfix padded batch by @ArtyomVorobev in #101

New Contributors

• @FuTSy13 made their first contribution in #99

Full Changelog: v0.5.1...v0.5.2

v0.5.1

What's Changed

• fixed cpc import by @ArtyomVorobev in #90
• add softmaxloss and tests by @ArtyomVorobev in #87
• MLM NSP Module by @mazitovs in #88
• fix test dropout error by @ivkireev86 in #91

New Contributors

• @ArtyomVorobev made their first contribution in #90
• @mazitovs made their first contribution in #88

Full Changelog: v0.5.0...v0.5.1

v0.5.0

What's Changed

• Fix metrics reset by @ivkireev86 in #72
• Pandas preprocessing without df copy, faster preprocessing for large datasets by @ivkireev86 in #73
• fix in supervised-sequence-to-target.ipynb by @blinovpd in #74
• ptls.nn.PBDropout by @ivkireev86 in #75
• tanh for rnn starter by @ivkireev86 in #76
• Auc regr metric by @ikretus in #78
• spatial dropout for NoisyEmbedding, LastMaxAvgEncoder, warning for bidir RnnEncoder by @justalge in #80
• Hparam tuning demo. hydra, optuna, tensorboard by @ivkireev86 in #81
• tabformer by @justalge in #83
• Supervised Coles Module, trx_encoder refactoring by @ivkireev86 in #84

New Contributors

• @blinovpd made their first contribution in #74

Full Changelog: v0.4.0...v0.5.0

v0.4.0

What's Changed

• Seq encoder refactoring by @ivkireev86 in #29
• regr.task ZILNLoss, RMSE, BucketAccuracy by @ikretus in #36
• lighting modules and nn layers refactoring by @ivkireev86 in #34
• Demo colab by @ivkireev86 in #40
• Fix drop target arrays by @ivkireev86 in #42
• feature naming by @ivkireev86 in #43
• Update abs_module.py by @justalge in #37
• Extended inference demo by @ivkireev86 in #45
• fix import path by @ivkireev86 in #46
• Experiments sync by @ivkireev86 in #50
• Experiments sync by @ivkireev86 in #52
• Target dist by @ikretus in #58
• Data load refactoring by @ivkireev86 in #60
• doc update by @ivkireev86 in #62
• doc update by @ivkireev86 in #63

New Contributors

• @ikretus made their first contribution in #36

Full Changelog: v0.3.0...v0.4.0

What's Changed

• Seq encoder refactoring by @ivkireev86 in #29
• regr.task ZILNLoss, RMSE, BucketAccuracy by @ikretus in #36
• lighting modules and nn layers refactoring by @ivkireev86 in #34
• Demo colab by @ivkireev86 in #40
• Fix drop target arrays by @ivkireev86 in #42
• feature naming by @ivkireev86 in #43
• Update abs_module.py by @justalge in #37
• Extended inference demo by @ivkireev86 in #45
• fix import path by @ivkireev86 in #46
• Experiments sync by @ivkireev86 in #50
• Experiments sync by @ivkireev86 in #52
• Target dist by @ikretus in #58
• Data load refactoring by @ivkireev86 in #60
• doc update by @ivkireev86 in #62
• doc update by @ivkireev86 in #63

New Contributors

• @ikretus made their first contribution in #36

Full Changelog: v0.3.0...v0.4.0

What's Changed

• Seq encoder refactoring by @ivkireev86 in #29
• regr.task ZILNLoss, RMSE, BucketAccuracy by @ikretus in #36
• lighting modules and nn layers refactoring by @ivkireev86 in #34
• Demo colab by @ivkireev86 in #40
• Fix drop target arrays by @ivkireev86 in #42
• feature naming by @ivkireev86 in #43
• Update abs_module.py by @justalge in #37
• Extended inference demo by @ivkireev86 in #45
• fix import path by @ivkireev86 in #46
• Experiments sync by @ivkireev86 in #50
• Experiments sync by @ivkireev86 in #52
• Target dist by @ikretus in #58
• Data load refactoring by @ivkireev86 in #60
• doc update by @ivkireev86 in #62
• doc update by @ivkireev86 in #63

New Contributors

• @ikretus made their first contribution in #36

Full Changelog: v0.3.0...v0.4.0

v0.3.0

More Pythonic Core API: constructor arguments instead of config objects

What's Changed

• cpc params by @justalge in #9
• All modules by @justalge in #15
• Mlm pretrain by @ivkireev86 in #13
• all encoders and get rid of get_loss by @justalge in #19
• init by @justalge in #20
• Documentation index by @ivkireev86 in #8
• Demos api update by @ivkireev86 in #18
• loss output correction by @ivkireev86 in #22
• Test fixes by @ivkireev86 in #23
• readme_demo_link by @ivkireev86 in #25
• init by @justalge in #26
• work without logger by @justalge in #7
• trx_encoder refactoring by @ivkireev86 in #28

Full Changelog: v0.1.2...v0.3.0