jiahaozhang2003/beacon-secondary-structure

BEACON — Secondarystructureprediction

RNA secondary-structure prediction data with nucleotide-level pair matrices.

Official data from the shared BEACON/RNABenchmark Drive folder: https://drive.google.com/drive/folders/19ddrwI8ycvIxkgSV3gDo_VunLofYd4-6?hl=en.

This repository is the standardized Hugging Face publication of the official task data. The data/ directory is the canonical viewer-friendly layer, and the original file contents and source names are preserved for reproducibility. For SSP, .npy targets are deterministically sharded into subdirectories to comply with the Hugging Face 10,000-files-per-directory limit; the original unsharded layout remains available in the GPUHub raw staging area. beacon_manifest.json records source file IDs, checksums, sizes, split counts, schema, and processing notes. Interrupted-upload artifacts ending in .partial are retained only in the GPUHub raw staging area and excluded from the published repository.

Canonical splits

Standardized schema

Standardization notes

• —Canonical split names are train, validation, and test; source val.csv/val_1.json is published as validation.
• —The official raw directory tree is preserved alongside the standardized data/*.parquet files.
• —Source file names and target paths are retained in beacon_manifest.json for provenance.
• —Parquet rows retain sequence, dot-bracket annotation, sequence length, and the raw pair-matrix target path.
• —SSP .npy targets are deterministically sharded under bpRNA/{TR0,VL0,TS0}/shard-XX to satisfy the Hugging Face 10,000-files-per-directory limit; source filenames and contents are unchanged.

Usage

from datasets import load_dataset

dataset = load_dataset("jiahaozhang2003/beacon-secondary-structure")

Reproducibility

This repository was generated on GPUHub from the public Drive source with the RiboLoop downstream benchmark data-staging script. No model outputs or training artifacts are included.