datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
GH114_activity_DPO
GH114 Enzyme Activity Dataset (from FLOP Benchmark)
Dataset Description
This dataset contains protein sequences, SaProt structure-aware sequence, and activity labels for 55 Glycoside Hydrolase Family 114 (GH114) enzymes.
It is specifically formatted for aligning the NorseDrunkenSailor/Qwen_smol_GH114 base model. This model is a generative Causal Language Model (CLM) adapted from Qwen2 and trained on SaProt 3Di-aware sequences. The goal is to use Direct Preference… See the full description on the dataset page: https://huggingface.co/datasets/NorseDrunkenSailor/GH114_activity_DPO.zsx_DPO_mainzsx_DPO_rank