All pull requests

Pull requests

• Add an MLX backend for TabFM v1.0.0 (Apple silicon)

  #106·franz101 opened Sep 22, 2026·

  ·
  1

• Bump anyio from 4.14.1 to 4.14.2

  dependenciespython
  #105·dependabot[bot] opened Sep 18, 2026Bot·

  ·

• Validate ensemble parameters before estimator fitting

  #104·ayoubchikri opened Sep 15, 2026·

  ·
  2

• Validate PyTorch transformer configuration

  #103·sylvesterkaczmarek opened Sep 10, 2026·

  ·
  1

  1

• Validate softmax temperature

  #102·sylvesterkaczmarek opened Sep 8, 2026·

  ·
  1

  1

• docs: document context caching options in README examples

  #101·astonishedrobo opened Sep 4, 2026Contributor·

  ·
  1

  3

• Fix classifier OOF validation slicing

  #98·paramp-ratap opened Aug 19, 2026·

  ·

• Fix #93: Reject non-positive softmax temperatures

  #96·Echo-Nie opened Aug 17, 2026Contributor·

  ·
  1

  1

• Resolve the Hugging Face token from HF_TOKEN or a stored login

  #87·kobihikri opened Aug 9, 2026·

  ·
  1

• Parameter Name Correction

  #86·rishiraj opened Aug 3, 2026·

  ·

• Bump setuptools from 81.0.0 to 83.0.0

  dependenciespython
  #84·dependabot[bot] opened Jul 25, 2026Bot·

  ·

• Add TabFMDataGenerator for synthetic tabular data generation

  #83·LckyLke opened Jul 24, 2026·

  ·
  1

  2

• Add sequence-parallel multi-GPU inference (PyTorch and JAX) and a cuDNN flash attention option

  #81·weihaokong opened Jul 23, 2026Collaborator·

  ·

• Rewrite prefix key-padding masks as k/v truncation to keep SDPA on the flash path (PyTorch)

  #80·arnaudlvq opened Jul 21, 2026·

  ·
  1

• README: note that the pre-trained weights are non-commercial (#29)

  #79·p4vlos opened Jul 21, 2026·

  ·
  1

  1

• Fix open bugs: MPS float64 crash (#68), OOF calibration with max_num_rows (#55), sklearn transform_output leak (#58), safetensors dep (#56)

  #78·p4vlos opened Jul 21, 2026·

  ·
  1

• Fix encoded_missing_value being ignored and all-missing datetime columns crashing

  #76·devYRPauli opened Jul 20, 2026Contributor·

  ·
  1

• Add safetensors to the pytorch extra

  #75·manjunathshiva opened Jul 20, 2026·

  ·

• Cast float64 targets before the device move in the context-cache path

  #74·manjunathshiva opened Jul 20, 2026·

  ·

• fix remove extra import

  #71·direkkakkar319-ops opened Jul 18, 2026Contributor·

  ·

• Cast float64 targets before moving them to the device

  #69·devYRPauli opened Jul 16, 2026Contributor·

  ·
  1

  2

• Add a README note about the TabFM technical report status

  #67·MayVerse4 opened Jul 11, 2026Contributor·

  ·
  1

• Expose prefill/decode context caching via fit_mode="fit_with_cache" (JAX)

  #61·jjovalle99 opened Jul 8, 2026·

  ·

• Add LoRA fine-tuning to the MLX backend

  #52·ovuruska opened Jul 4, 2026·

  ·

• Add MLX backend for Apple silicon

  #49·ovuruska opened Jul 4, 2026·4 tasks done·

  ·
  1