Dataset
kl3m-data-edgar-10-q
The KL3M Data Project from the ALEA Institute supplies copyright-clean training material, and the dataset listed here is one component pending further documentation on its dedicated page.
Data preview
Dataset card
The ALEA Institute KL3M EDGAR 10-Q dataset is a copyright-clean release of SEC Form 10-Q quarterly filings, distributed as part of the KL3M Data Project for training large language models. It packages standardized text representations of these filings in a single Parquet split.
What this dataset contains
The dataset holds roughly 1.5 million pre-processed 10-Q filings sourced from SEC EDGAR, delivered as one Parquet file under the train split. Each row represents a single filing document retrieved from the S3 bucket at data.kl3m.ai/documents/edgar/, stored in text/markdown format. The collection is sized between one million and ten million rows according to its Hugging Face size tag.
Structure and fields
Four columns are exposed in the schema: identifier (string), dataset (string), mime_type (string), and tokens (sequence of int64). The identifier field carries an S3-style path pointing to the source JSON document, the dataset field is uniformly populated with the value "edgar", and mime_type records the content format as text/markdown. The tokens column contains the pre-tokenized integer sequence used directly for language model training.
How analysts use it
Because each row preserves the full text of a public-company quarterly filing, analysts can mine the corpus for boilerplate language trends, disclosure changes, and management sentiment across reporting periods. The pre-tokenized tokens column also allows direct fine-tuning of domain-specific language models on financial filing text, supporting automated summarization of 10-Q disclosures. The standardized markdown format and open CC-BY-style release simplify reproducible research across institutions.
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.