datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
soxdata_encodec
Dataset Card for "soxdata_encodec"
More Information needed
sox_speech_tokenizer
Dataset Card for "sox_speech_tokenizer"
More Information needed
aliafzal9323_soxx-ishares-semiconductor-etf-daily-2001-2026
SOXX iShares Semiconductor ETF Daily (2001-2026)
Daily OHLCV price data for iShares Semiconductor ETF (SOXX) spanning 24+ years
Dataset Info
Source: Kaggle
Original Size: 0.11 MB
Kaggle Downloads: 6
Files: 1
Files
SOXX_Daily_Stock_Data.csv
Mirrored from Kaggle
Soxcrtcd
Dataset Card for OAM-TCD: A globally diverse dataset of high-resolution tree cover maps
Annotation example in OAM-TCD (ID 1445), RGB image licensed CC BY-4.0, attribution contributors of OIN.
Left: RGB aerial image, Middle: annotations shown, distinguished by instance ID, Right: annotations identified by class (blue = tree, orange = canopy)
Dataset Details
OAM-TCD is a dataset of high-resolution (10 cm/px) tree cover maps with instance-level masks for 280k trees and… See the full description on the dataset page: https://huggingface.co/datasets/soxnjeiaka/tcd.ardb-layout-coco-v1
ARDB Daily Bulletin Layout Detection (COCO, folder format)
⚠️ Please use Soxavin/ardb-layout-coco-v2 instead
This v1 repository is the original folder / COCO export and is no longer maintained.
The identical data is published in Parquet format — with a working Dataset Viewer and the
standard 🤗 datasets loader — at
Soxavin/ardb-layout-coco-v2.
New users should load v2:
from datasets import load_dataset
ds = load_dataset("Soxavin/ardb-layout-coco-v2")
A… See the full description on the dataset page: https://huggingface.co/datasets/Soxavin/ardb-layout-coco-v1.ardb-sft-v5
ARDB Unified Page Understanding (bbox + text, one forward pass)
Vision fine-tuning data for document-understanding VLMs, built from ARDB (Agricultural and
Rural Development Bank of Cambodia) daily market-price bulletins. This dataset unifies layout
detection (bbox only) and transcription (text only, on pre-cropped regions): one full page
image in, one JSON list out, where each region carries its box AND its transcribed text
together, in a single forward pass.
Model-agnostic: the… See the full description on the dataset page: https://huggingface.co/datasets/Soxavin/ardb-sft-v5.ardb-gemma-sft-v2
Note: A newer, larger version of this dataset lineage is available at
Soxavin/ardb-sft-v5 — same task
design (one page image in, boxes + text out in a single forward pass), expanded from a
single-year corpus (28 documents) to a multi-year one (47 documents, 2022–2026), with an
era-stratified train/validation/test split. This version is kept as-is for reproducibility
of results computed against these exact 66/9/9 rows; use v5 for new work.
ARDB Unified Page Understanding (bbox +… See the full description on the dataset page: https://huggingface.co/datasets/Soxavin/ardb-gemma-sft-v2.ardb-layout-coco-v2
Note: A newer, larger version of this dataset is available at
Soxavin/ardb-layout-coco-v5
— same task and schema, expanded from a single-year corpus (30 documents) to a multi-year
one (55 documents, 2022–2026), with adjacent-day leakage protection and era-stratified
splits. This version is kept as-is for reproducibility of prior results; use v5 for new work.
ARDB Daily Bulletin Layout Detection (COCO)
A document-layout object-detection dataset built from ARDB (Agricultural and… See the full description on the dataset page: https://huggingface.co/datasets/Soxavin/ardb-layout-coco-v2.ardb-layout-coco-v5
ARDB Daily Bulletin Layout Detection (COCO)
A document-layout object-detection dataset built from ARDB (Agricultural and Rural
Development Bank of Cambodia) daily market-price bulletins — born-digital Khmer-language PDFs,
one price table per page. Every page is annotated with bounding boxes for five layout regions
(table, text, section header, page furniture, picture).
Each row is a full page image with its box annotations embedded, so the Dataset Viewer shows
the image beside… See the full description on the dataset page: https://huggingface.co/datasets/Soxavin/ardb-layout-coco-v5.sox-itgc-jml-synthetic-dataset
Latest Additions
This repository now includes two synthetic assets:
JML Leaver Evidence Dataset
500 synthetic SOX ITGC JML leaver access testing cases
Labels: pass, fail, review_required
Splits: train, val, test
Includes evidence summaries, extracted facts, exception types, and audit reasoning
Synthetic Company Policy Dataset
80 synthetic access termination and deprovisioning policies
Includes standard access SLA, privileged access SLA, weekend/business-day rules, required… See the full description on the dataset page: https://huggingface.co/datasets/auditbrain/sox-itgc-jml-synthetic-dataset.my-covid-datasetsoxdata_encodec
Dataset Card for "soxdata_encodec"
More Information needed