mirror mix-phase1-40M token shards from S3

mirror mix-v2-40M token shards from S3 (part 5)

mirror mix-v2-40M token shards from S3 (part 4)

mirror mix-v2-40M token shards from S3 (part 3)

mirror mix-v2-40M token shards from S3 (part 2)

mirror mix-v2-40M token shards from S3

docs: top-level README + mixes/ (catalog, per-mix configs, mix_dataset.py builder)

Converted to Selfies & Parquet

Fix script for strings beyond first whitespace

Splitting large SMILES files @1GB size

forgot strict

batched loading

Selfies_converter added

gdb added and reformatted others

turned csv from Kaggle into smi

move Kaggle

Kaggle Datasets

Emolecules & mcule_full

initial commit