datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
toy-models-of-sft-data
Toy Models of SFT Data
This is a public-clean candidate data package for the Toy Models of SFT project.
It is built for researcher inspection first.
The package answers two questions:
What were the models trained on?
How did the models actually behave under evaluation?
The package includes training data, eval inputs, model rollouts, judge scores,
parsed GPQA outputs, aggregate tables, paper figures, frozen plot data, and
provenance records. It deliberately includes some… See the full description on the dataset page: https://huggingface.co/datasets/matonski/toy-models-of-sft-data.retinoblastomaRetinoblastoma Dataset
This dataset contains information related to retinoblastoma from ClinvarTuring https://github.com/ToyokoLabs/ClinvarTuring
Licensing Information
License: cc-by-4.0
Authors
Morgan Lyu, Sebastian Bassi and Virginia Gonzalez
rag-observatory-toy-traces
RAG Observatory Toy Traces
Three small, synthetic traces for testing RAG diagnostics and report interfaces.
Each example isolates a different outcome:
a supported answer with one irrelevant retrieved document;
a retrieval miss that sends the wrong evidence to the generator;
an answer that contradicts relevant selected context.
The records mirror the examples used by
GioiaZheng/rag-observatory
and its interactive Space.
Intended use
This dataset is suitable for:… See the full description on the dataset page: https://huggingface.co/datasets/GioiaZheng/rag-observatory-toy-traces.ODE-Toy-RL-Data
ODE Data
Dataset layout:
train/metadata.jsonl: one row per example with question, answer, and image metadata.
train/*.{jpg,png,webp}: decoded images referenced by file_name in metadata.
raw_training_format/train_verl.parquet: training-compatible parquet format for the same examples.
Each example has exactly one image. The public viewer split intentionally omits source id fields.