Dataset
zephyr_pi0_gen_57k_for_offline_dpo_ipo
Dataset Card for "zephyr_pi0_gen_57k_for_offline_dpo_ipo" More Information needed
Data preview
Dataset card
zephyr_pi0_gen_57k_for_offline_dpo_ipo is a small text corpus published by user raftrsf, distributed as a parquet file with train and test splits totaling roughly 57,464 rows. The publisher has not supplied a license, descriptive readme, or stated dataset size beyond its size band.
What this dataset contains
The dataset is delivered as a parquet file tagged with the text modality and is intended for offline direct preference optimization workflows, as suggested by the name. It carries the size_categories tag indicating between 10,000 and 100,000 rows, with 57,464 rows split between train and test. The publisher has not documented the topical scope, and the readme description only points to a missing more-information section.
Structure and fields
Two columns are exposed: chosen and rejected, each represented as arrays of objects where each object contains a string content field and a role field. The columns data type is nested, with content typed as string Value. Both columns are present on every row, providing paired preferred and rejected responses typical of preference training data.
How analysts use it
Because the schema offers only paired chosen and rejected text fields with no business identifiers, securities coverage, or financial metadata, no concrete investment-fund use case can be supported by the published facts. The only plausible application is loading the file with the pandas, polars, datasets, or mlcroissant libraries to fine-tune or evaluate a small language model on the preference pairs. Anything beyond that, including any market or issuer signal, would require information the publisher has not disclosed.
Documentation & references 1
Related datasets
Browse moreParsed eligibility criteria from 13,229 ClinicalTrials.gov studies representing the candidate pool gathered by two first-stage retrievers during TREC Clinical Trials 2021–2023, formatted as typed entity-relation graphs for reranking models.
The clinical-trials-patient-graphs dataset is a small tabular collection of patient-level clinical records distributed in Parquet format by the 2001jdev publisher. It organizes entities and relations extracted from trial data into structured rows spanning 2021 through 2023.
This dataset is a parsed snapshot of ClinicalTrials.gov records, distributed in Parquet format on the Hugging Face Hub under an unspecified license. It contains roughly 52,000 trial entries drawn from public registry filings.
clinical-trials-trec-qrels is a tabular relevance-judgment file distributed by the 2001jdev user on the Hugging Face Hub. It maps clinical-trial topic identifiers to NCT registry IDs with graded relevance scores.
The clinical-trials-trec-topics dataset is a small tabular collection of clinical-trial topic records distributed in Parquet format on the Hugging Face Hub. It appears to be derived from TREC clinical-trial retrieval topics, with rows partitioned by year.
SEC EDGAR 10-K & 10-Q financial statements for 2,600+ NASDAQ companies.
Discussion
Questions or observations about this dataset? Join the conversation on X.
No comments yet. Start the conversation on X.