All issues

JAX and PyTorch constructors disagree on identical config: same value builds on one backend, raises on the other

#95

· kobihikri opened

on Aug 9, 2026

Estimators define no _parameter_constraints; all 13 numeric parameters accept 0 and negative values

#94

· kobihikri opened

on Aug 9, 2026

Negative softmax_temperature silently inverts predictions; zero yields all-nan probabilities

#93

· kobihikri opened

on Aug 9, 2026

TabFM constructor accepts zero/negative dimensions; col_num_blocks=0 builds a model with no attention blocks

#92

· kobihikri opened

on Aug 9, 2026

config.json keys are merged into the model constructor unvalidated (OOM kill, ZeroDivisionError)

#91

· kobihikri opened

on Aug 9, 2026

PyTorch backend silently accepts shapes/activations the JAX backend rejects (three cases)

#90

· kobihikri opened

on Aug 9, 2026

JAX RotaryEmbedding: assert admits dim==2, then divides by (dim-2) -> ZeroDivisionError

#89

· kobihikri opened

on Aug 9, 2026

Pin the Hugging Face revision when loading pretrained weights

#88

· kobihikri opened

on Aug 9, 2026

Support for synthetic data generation

#82

· LckyLke opened

on Jul 24, 2026

PyPI release (1.0.0) is incompatible with the published Hugging Face checkpoint(for kaggle); loader fix exists on main but has not been released

#77

· Samarth-ITM opened

on Jul 20, 2026

PyTorch backend: crash on Apple MPS with float64 targets — device move precedes the float64->float32 guard

#68

· phas02 opened

on Jul 13, 2026

Request for the Technical Report of TabFM

#64

· theaimer09 opened

on Jul 9, 2026