Repository navigation

Fix PubChemBatched dataset - #170

Merged

sfluegel05 merged 6 commits into

devfrom

fix/pubchem-batched

Jun 26, 2026

Commits on May 28, 2026
- fix get_data_splits for pubchem batched
  sfluegel05
  committed
  e42c705
- overwrite load_processed_data to reload batch for every epoch
  sfluegel05
  committed
  ce5ffde
Commits on May 29, 2026
- add hard cutoff for feature length
  sfluegel05
  committed
  f41319f
Commits on Jun 1, 2026
- add process_batch for electra pre
  sfluegel05
  committed
  9660cee
- reshape attention mask
  sfluegel05
  committed
  92650f7
- adjust default vocab_size to static tokens
  sfluegel05
  committed
  418c5ee