Add AudioSkillsXL and F0Bioacoustic datasets#259
Open
david-rx wants to merge 1 commit into
Open
Conversation
AudioSkillsXL: Multi-task audio understanding dataset combining conversation-style QA from nvidia/AudioSkills with AudioSet audio. Six sources (counting_qa, wavcaps, fsd50k, clotho_v2, audioset, audioset_sl) loadable individually or as a union via the "all" split. Includes structured messages column and pre-resampled audio paths. F0Bioacoustic: ~250k non-human vocalizations from 13 taxa with ground-truth fundamental frequency contours. Supports taxa filtering, pre-resampled 16k/32k audio, and F0 contour parsing into DataFrames. Both datasets follow existing esp-data conventions with full test suites (37 tests total). Made-with: Cursor
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
AudioSkillsXL: Multi-task audio understanding dataset combining conversation-style QA from nvidia/AudioSkills with AudioSet audio. Six sources (
counting_qa,wavcaps,fsd50k,clotho_v2,audioset,audioset_sl) loadable individually or as a union via the"all"split. Handles structuredmessagescolumn (JSON chat turns), dual data roots for local vs AudioSet-backed audio, and pre-resampled 16k/32k audio paths.F0Bioacoustic: ~250k non-human vocalizations from 13 taxa with ground-truth F0 contours (Musikhin et al. 2025). Supports taxa filtering, pre-resampled 16k/32k audio, inline TSV F0 contour parsing into DataFrames, and Nyquist flagging. CC0-1.0 licensed.
Both follow existing esp-data conventions:
@register_dataset,DatasetInfo,from_config, full numpy-style docstrings,_process/__getitem__/__iter__protocol.Test plan
test_audio_skills_xl.py(17 tests): info, data, columns, splits, sources, length, getitem, iteration, invalid split/sources, consistency, transforms from config, output mapping, audio processing, str repr, index error, messages parsingtest_f0_bioacoustic.py(20 tests): info, data, columns, splits, taxa, length, getitem, F0 contour parsing, iteration, invalid split/taxa, taxa filtering, consistency, transforms from config, output mapping, data root, audio processing, str repr, index error, get_available_labelsMade with Cursor