Portfolio
2023-2025Datasets

Training Datasets

Datasets for Training, Reasoning, and RLHF

5 Public Datasets · Powering #1 Models & RLHF Experiments

Custom datasets built for targeted training experiments. The simple-math family explores minimal arithmetic corpora for reasoning under SFT and DPO. Tree of Knowledge introduces symbolic knowledge structuring. The german-humanlike pair demonstrates downstream artifacts produced by composing SingleMoM RLHF experts.

Client
Independent — Juanako.AI
Role
Dataset author
Duration
2023-2025
Team
Solo
Outcomes
5
Public Datasets
800K rows
Largest
Yes
Used in #1 Models
Math · Knowledge · Style
Coverage
Datasets
NameDateSizeTypePurpose / Used In
simple-math20-Jan-2024779K rowsSFTArithmetic reasoning · SimpleSmaug #1 34B
simple-math-DPO27-Jan-2024800K rowsDPOArithmetic preference training
Tree of Knowledge24-May-2023~5MBSymbolicKnowledge structure · Cybertron 7B v1/v2
german-humanlike-clean-1k26-Mar-2025856 rowsRLHFSingleMoM expert composition (curated)
german-humanlike-large26-Mar-202510.8K rowsRLHFSingleMoM expert composition (full)
datasethuggingfacesynthetic-datarlhfdposftdata-engineering