HauserGroup/ApeTokenizer-SMILES

1{2  "ape_source": "modernmolbert.local",3  "created_at_utc": "2026-05-22T04:05:53.431998+00:00",4  "creation_command": "python -m modernmolbert.train_ape_tokenizer",5  "dataset_name": "data/pretrain/chembl36_selfies",6  "extra_vocab_selfies_path": null,7  "extra_vocab_symbols_added": 0,8  "extra_vocab_symbols_path": null,9  "extra_vocab_symbols_requested": 0,10  "max_merge_pieces": 6,11  "max_vocab_size": 2000,12  "min_freq_for_merge": 3000,13  "molecule_column": "smiles_canonical_clean",14  "representation": "SMILES",15  "seed": 42,16  "shuffle_buffer_size": 100000,17  "special_ids": {18    "bos_token": 0,19    "eos_token": 2,20    "mask_token": 4,21    "pad_token": 1,22    "unk_token": 323  },24  "tokenizer_path": "tokenizer/chembl36_smiles_2m_ape_max6_mf3000.json",25  "tokenizer_sha256": "faf7748e8959b252c9d0ad83c2228df37a45dc9a68c15ead1ced2942cc8f155e",26  "tokenizer_train_size": 2000000,27  "vocab_size": 138628}29