Code

Releasing Version 1.0.0#172

Code

Merged

AnFreTh merged 153 commits intoreleasebasf/mamba-tabular:releasefrom masterbasf/mamba-tabular:masterCopy head branch name to clipboardDec 4, 2024

+9,047-1,949

Commits

Commits on Sep 4, 2024

• include MambAttn

  AnFreTh

  committed
  9998d38

• adapt mabattn config

  AnFreTh

  committed
  de0886e

Commits on Sep 6, 2024

• pruning for hpo

  AnFreTh

  committed
  d39c056

• include config mapper for hpo

  AnFreTh

  committed
  c1467bd

• include gp_minimize in sklearn base regressor

  AnFreTh

  committed
  85adc31

Commits on Sep 11, 2024

• fix bug in TabTransformer embedding_layer

  AnFreTh

  committed
  6a6a51b

• mlp basemodel convenience fix

  AnFreTh

  committed
  29811d6

• resnet convenience fix

  AnFreTh

  committed
  1cb93c5

Commits on Sep 13, 2024

• config convenience fix

  AnFreTh

  committed
  4ef4b5d

• add efficiency scripts

  AnFreTh

  committed
  bf90683

• remove citation for anonymity

  AnFreTh

  committed
  2c4bcb8

Commits on Sep 28, 2024

• add hpo to classifier and lss

  AnFreTh

  committed
  797b624

• minor pooling error in mambular

  AnFreTh

  committed
  5b05664

• add conv layer to rnn for positional invariance

  AnFreTh

  committed
  14663e0

• add convrnn to base class

  AnFreTh

  committed
  4bd83e9

• adjust config of RNN

  AnFreTh

  committed
  b3327a1

• include optimizer args in taskmodel

  AnFreTh

  committed
  3d76805

• adapt sklearn classes to allow optimizer kwargs

  AnFreTh

  committed
  81ae17f

• adjust default optimizer

  AnFreTh

  committed
  bce26cb

• include skopt in requirements

  AnFreTh

  committed
  28d56b6

• Merge pull request #138 from basf/attn

  Show description for ed5a0f3
  AnFreTh

  authored
  ed5a0f3

Commits on Oct 21, 2024

• include pscan and original mamba_ssm triton version

  AnFreTh

  committed
  4723942

• fix import

  AnFreTh

  committed
  df60c1c

• fix import and config

  AnFreTh

  committed
  85a468b

• adjust attn_config

  AnFreTh

  committed
  3fdfa96

• fix d_interemediate-d_state error

  AnFreTh

  committed
  949bdb1

• input of original ssm forward pass fix

  AnFreTh

  committed
  161ceeb

• forward pass original mamba tryout

  AnFreTh

  committed
  141b76f

• create ResidualBlock for Original Mamba-ssm

  AnFreTh

  committed
  c0322bc

• fix residualblock **kwargs

  AnFreTh

  committed
  0dd1b37

• fix RMSNorm init

  AnFreTh

  committed
  4b9066e

Commits on Oct 24, 2024

• try-out Mamba2

  AnFreTh

  committed
  0f57b26

• adjust config to new params

  AnFreTh

  committed
  c10d40a

• adjust mamba-ssm import

  AnFreTh

  committed
  c90c9fd

• adjust _lazy_import in ResidualBlock

  AnFreTh

  committed
  b424e8b

• include configs in package

  AnFreTh

  committed
  9a04905

• adjust mambatab config to use mamba-ssm and mamba2

  AnFreTh

  committed
  31ddf22

• adjust docstrings to document new params

  AnFreTh

  committed
  8b08f8e

• delete unnecessary imports in mambatab

  AnFreTh

  committed
  426e6c6

• include new models in readme

  AnFreTh

  committed
  617d6f3

• include mamba-ssm installation in readme

  AnFreTh

  committed
  10881eb

• Merge pull request #145 from basf/efficiency

  Show description for 95e87dd
  AnFreTh

  authored
  95e87dd

Commits on Nov 5, 2024

• fix preprocessor kwargs typo

  AnFreTh

  committed
  ef04001

• restructure utils and add neural decision tree

  AnFreTh

  committed
  1c529f0

• Adjust new imports

  AnFreTh

  committed
  de77ed5

• add neural Decision Forest base architecture

  AnFreTh

  committed
  7322839

• add ndtf to new models in __init__

  AnFreTh

  committed
  3f25fd6

• add new configs. include mLSTM/sLSTM in rnn config

  AnFreTh

  committed
  385f2dd

• Merge pull request #149 from basf/NDTF_LSTM

  Show description for 54ca398
  AnFreTh

  authored
  54ca398

• add ntdf config in init

  AnFreTh

  committed
  bdac22b

• add sparsemax

  AnFreTh

  committed
  9eb5d42

• data-aware initialization module

  AnFreTh

  committed
  ab3abbf

• utils func for checking if tensor or np.array

  AnFreTh

  committed
  473db6b

• add ODST and DenseBlock

  AnFreTh

  committed
  e59bfcb

• add node into basemodels - includes tabular MLP head

  AnFreTh

  committed
  8d362df

• add default config for NODE model

  AnFreTh

  committed
  8b12c61

• add Node to models and __init__

  AnFreTh

  committed
  90e1476

• refactor normalization layer -> get_normalization_layer included in __init__

  AnFreTh

  committed
  9bdbff3

• add nodeconfig in __init__

  AnFreTh

  committed
  b0c0bf4

• fix typo in docstrings

  AnFreTh

  committed
  bea4bc3

• Merge pull request #151 from basf/NODE

  Show description for 61bb9b3
  AnFreTh

  authored
  61bb9b3

• adapt config for normalization layer in rnn

  AnFreTh

  committed
  4e3bcda

• Merge pull request #152 from basf/rnn_fix

  Show description for bb11e81
  AnFreTh

  authored
  bb11e81

Commits on Nov 11, 2024

• adjust readme and include new models

  AnFreTh

  committed
  fb74d8e

• LinearBatchEnsemlbe layer as used in TabM paper

  AnFreTh

  committed
  ef1166d

• only use config in embedding layer as arg

  AnFreTh

  committed
  1bb2dc4

• allow for None as input

  AnFreTh

  committed
  5998fd3

• rename MLp to MLPhead and only use config as input

  AnFreTh

  committed
  abf741d

• use config as input in ConvRNN and introduce batchEnsemble RNN layer

  AnFreTh

  committed
  d0440bb

• only use config in TransformerEncoder Layer

  AnFreTh

  committed
  6fb11fa

• include pooling and init pooling in basemodel class

  AnFreTh

  committed
  fd74037

• new arch from utils -> only config as arg

  AnFreTh

  committed
  3212cc5

• adjust all models to new embeddinglayer and new layer utils

  AnFreTh

  committed
  230b276

• include TabM as introduce in paper

  AnFreTh

  committed
  d80d16d

• batch Ensemble RNN -> todo bidirectional

  AnFreTh

  committed
  a570166

• include tabm and batchtabrnn configs

  AnFreTh

  committed
  1ba1064

• delete bidirectional from config

  AnFreTh

  committed
  51d71d1

• add layer_norm_eps to config

  AnFreTh

  committed
  10dca1f

• include batchtabrnn for reg/class/lss

  AnFreTh

  committed
  b33b2aa

• new model

  AnFreTh

  committed
  38edb67

• remove default values for lr related params in fit

  AnFreTh

  committed
  bafcde3

• delete lr related default params in fit

  AnFreTh

  committed
  2414fa3

• lr realted param adjustments

  AnFreTh

  committed
  8091909

• Merge pull request #154 from basf/TabM

  Show description for 9a236a5
  AnFreTh

  authored
  9a236a5

• make usable even when params not in config

  AnFreTh

  committed
  ea184ce

• adapt embedding layer to plr encodings

  AnFreTh

  committed
  063b8dd

• PLR layer inclusion

  AnFreTh

  committed
  786e4d2

• minor fix in embedding layer creation

  AnFreTh

  committed
  b900b71

• adjust defaults

  AnFreTh

  committed
  0986c65

• include new models in init

  AnFreTh

  committed
  058bad9

• Merge pull request #155 from basf/TabM

  Show description for 4d1f787
  AnFreTh

  authored
  4d1f787

Commits on Nov 12, 2024

• fix validation dataset bug

  AnFreTh

  committed
  7de4982

• Merge pull request #156 from basf/val_data-fix

  Show description for 74ad3aa
  AnFreTh

  authored
  74ad3aa

• original_mamba dt_rank fix

  AnFreTh

  committed
  097c6f4

Commits on Nov 13, 2024

• Merge pull request #157 from basf/original_mamba-fix

  Show description for 254827c
  AnFreTh

  authored
  254827c

• including tab_mini

  AnFreTh

  committed
  9650d24

• fix multiple predictions in ensemble loss

  AnFreTh

  committed
  0cc9501

Commits on Nov 14, 2024

• adjust how values are retrieved from config

  AnFreTh

  committed
  2a9fc96

• save config as hparams

  AnFreTh

  committed
  8f3994b

• fix batch-mini rnn arch

  AnFreTh

  committed
  c823188

• adjust models to new self.hparams arch

  AnFreTh

  committed
  6ca631d

• include ensembling loss

  AnFreTh

  committed
  e8b123c

• adjust configs to new params

  AnFreTh

  committed
  98b873f

• sklearn classes for new ensembling logic

  AnFreTh

  committed
  3cfa7ba

• fix proba prediction for ensembles

  AnFreTh

  committed
  bab7f11

• add positional_invariance layer

  AnFreTh

  committed
  6a83537

Commits on Nov 18, 2024

• fix new mamba2 version

  AnFreTh

  committed
  2bb90f5

Commits on Nov 19, 2024

• Bump lxml-html-clean from 0.1.1 to 0.4.0 in /docs

  Show description for 025705d
  dependabot[bot]

  authored
  025705d

Commits on Nov 22, 2024

• include new ensemble method for rnn

  AnFreTh

  committed
  4d82e0e

• adapt readme to new models

  AnFreTh

  committed
  a791bea

• adjust default parameters in configs for hpo

  AnFreTh

  committed
  fe632ab

Commits on Nov 28, 2024

• adjust categorical embedding

  AnFreTh

  committed
  c662fca

• make norm in resnet to layernorm for hpo

  AnFreTh

  committed
  f0f7c9b

• add ensemble configs

  AnFreTh

  committed
  8a8e9be

• adjust configs for better readibility

  AnFreTh

  committed
  9722342

• adding attention batch-ensemble

  AnFreTh

  committed
  f90eb1f

Commits on Nov 29, 2024

• Merge branch 'tab-mini' of https://github.com/basf/mamba-tabular into tab-mini

  AnFreTh

  committed
  6b2b2ac

Commits on Dec 2, 2024

• adding tabular cnn

  AnFreTh

  committed
  b1eccd6

• adding transformer BE layer

  AnFreTh

  committed
  650c911

• adjusting embedding layer

  AnFreTh

  committed
  3b8e585

• adjusting embedding layer

  AnFreTh

  committed
  27ff556

• including ftet

  AnFreTh

  committed
  578ff77

• adjusting default configs with new doc structure

  AnFreTh

  committed
  fe161b1

• including new models and adjusting sklearn classes to hpo

  AnFreTh

  committed
  4f91525

• add verbose to preprocessing

  AnFreTh

  committed
  fe3c1d7

• .

  AnFreTh

  committed
  c0a62ca

• adjust import

  AnFreTh

  committed
  3bbe703

• remove normalization layer

  AnFreTh

  committed
  0709050

• adjust mambular config to new params

  AnFreTh

  committed
  ca9dff0

• fix hpo bug

  AnFreTh

  committed
  6d5f843

• imrpove hpo config mapper defaults

  AnFreTh

  committed
  c6b266f

Commits on Dec 3, 2024

• adapt models to new preprocessor

  AnFreTh

  committed
  a35f5c0

• include new preprocessor funcs and adapt get_info logic

  AnFreTh

  committed
  f3c218d

• adjust embedding layer to new preprocessing

  AnFreTh

  committed
  112cd96

• delete model class

  AnFreTh

  committed
  142d4b2

• adjust configs

  AnFreTh

  committed
  627d48a

• include cat_preprocessing in preprocessor arg names

  AnFreTh

  committed
  a696987

• include float preprocessing

  AnFreTh

  committed
  c354c00

• Merge pull request #163 from basf/tab-mini

  Show description for 328bf8a
  AnFreTh

  authored
  328bf8a

• adapt readme

  AnFreTh

  committed
  eadafe1

• Merge pull request #164 from basf/temp_develop

  Show description for 1ee79ea
  AnFreTh

  authored
  1ee79ea

• adapt paper link in readme

  AnFreTh

  committed
  cf5f64b

• Merge pull request #165 from basf/hot-fix_develop

  Show description for e5c2892
  AnFreTh

  authored
  e5c2892

• Merge pull request #159 from basf/dependabot/pip/docs/lxml-html-clean-0.4.0

  Show description for d631275
  AnFreTh

  authored
  d631275

Commits on Dec 4, 2024

• add prepro-args to sklearn hpo

  AnFreTh

  committed
  99a309b

• Merge pull request #166 from basf/gridsearch

  Show description for 978c49e
  AnFreTh

  authored
  978c49e

• version bump

  AnFreTh

  committed
  4e4cde8

• Merge pull request #167 from basf/version_bump

  Show description for a4b624c
  AnFreTh

  authored
  a4b624c

• add ndtf in readme

  AnFreTh

  committed
  64417d1

• Merge pull request #168 from basf/ndtf-readme-fix

  Show description for d127835
  AnFreTh

  authored
  d127835

• Resolved conflicts between master and develop

  AnFreTh

  committed
  872404c

• Merge pull request #170 from basf/resolved-develop

  Show description for 7e22659
  AnFreTh

  authored
  7e22659

• Merge pull request #171 from basf/develop

  Show description for 0a829f6
  AnFreTh

  authored
  0a829f6