Upload README.md with huggingface_hub

README.md

@@ -1,172 +1,121 @@
 ---
 license: cc-by-4.0
 task_categories:
-- audio-classification
+  - audio-classification
+  - audio-to-audio
 tags:
-- drum-transcription
-- midi
-- velocity
-- electronic-drums
-- groove
+  - drum-transcription
+  - midi
+  - velocity
+  - electronic-drums
+  - groove
 pretty_name: Expanded Groove MIDI Dataset (E-GMD)
 size_categories:
-- 10K<n<100K
-dataset_info:
-  features:
-  - name: audio
-    dtype: audio
-  - name: filename
-    dtype: string
-  - name: drummer
-    dtype: string
-  - name: session
-    dtype: string
-  - name: bpm
-    dtype: float64
-  - name: beat_type
-    dtype: string
-  - name: time_signature
-    dtype: string
-  - name: duration
-    dtype: float64
-  - name: midi
-    dtype: binary
-  splits:
-  - name: train
-    num_bytes: 77861501194
-    num_examples: 35217
-  - name: validation
-    num_bytes: 5870763089
-    num_examples: 5031
-  - name: test
-    num_bytes: 17548778222
-    num_examples: 5289
-  download_size: 131987559338
-  dataset_size: 101281042505
+  - 10K<n<100K
 configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: validation
-    path: data/validation-*
-  - split: test
-    path: data/test-*
+  - config_name: raw
+    data_files:
+      - split: train
+        path: data/raw/train/*.tar
+      - split: validation
+        path: data/raw/validation/*.tar
+      - split: test
+        path: data/raw/test/*.tar
+  - config_name: features
+    data_files:
+      - split: train
+        path: data/features/train/*.tar
+      - split: validation
+        path: data/features/validation/*.tar
+      - split: test
+        path: data/features/test/*.tar
 ---
 
 # Expanded Groove MIDI Dataset (E-GMD)
 
-## Quick Start
+FLAC-compressed WebDataset mirror of the [E-GMD](https://magenta.tensorflow.org/datasets/e-gmd) dataset with precomputed drum transcription features.
 
-```python
-from datasets import load_dataset
-
-# Stream to avoid downloading the entire dataset
-ds = load_dataset("schismaudio/e-gmd", streaming=True)
-
-# Or download locally
-ds = load_dataset("schismaudio/e-gmd")
-```
-
-## Dataset Description
-
-The **Expanded Groove MIDI Dataset (E-GMD)** contains 444 hours of aligned MIDI and audio from 43 drum kits on a Roland TD-17 electronic drum kit. The dataset comprises 45,537 sequences derived from 1,059 unique performances by 10 drummers.
-
-E-GMD is an extension of the [Groove MIDI Dataset (GMD)](https://huggingface.co/datasets/schismaudio/groove-midi-dataset) that re-renders each original performance through 43 different electronic drum kits, providing massive timbral diversity while preserving the original expressive MIDI data. This makes E-GMD particularly valuable for training drum transcription models that need to generalize across different drum sounds.
-
-## Dataset Structure
-
-### Data Fields
-
-| Field | Type | Description |
-|-------|------|-------------|
-| `audio` | `Audio` | WAV audio file |
-| `filename` | `string` | Original filename |
-| `drummer` | `string` | Anonymized drummer ID |
-| `session` | `string` | Recording session identifier |
-| `bpm` | `float` | Tempo in beats per minute |
-| `beat_type` | `string` | "beat" (steady pattern) or "fill" (fill passage) |
-| `time_signature` | `string` | Time signature (e.g., 4-4, 6-8) |
-| `duration` | `float` | Duration in seconds |
-| `midi` | `bytes` | Raw MIDI data |
-
-### Data Splits
-
-| Split | Examples | Description |
-|-------|----------|-------------|
-| `train` | ~36,000 | Training set |
-| `validation` | ~5,000 | Validation set |
-| `test` | ~4,500 | Test set |
+## Dataset Overview
 
-Splits are partitioned by drummer to prevent data leakage — no drummer appears in more than one split.
+| Property | Value |
+|---|---|
+| Samples | 45,537 (WAV + MIDI pairs) |
+| Performances | 1,059 unique across 43 drum kits |
+| Drummers | 10 |
+| Audio | 16-bit FLAC, 44100 Hz, mono |
+| Splits | train (35,217) / validation (5,031) / test (5,289) |
 
-## Class Taxonomy
+## Configs
 
-The dataset uses raw MIDI pitches from the Roland TD-17 electronic drum kit. The 9 canonical drum classes from the original GMD taxonomy apply:
+| Config | Content | Est. Size |
+|---|---|---|
+| `raw` | FLAC audio + MIDI per sample | ~70 GB |
+| `features` | Mel specs (128xT) + onset targets (26xT) + velocity targets (26xT) | ~44 GB |
 
-| Class | Instrument | MIDI Pitches |
-|-------|-----------|-------------|
-| 0 | Kick | 36 |
-| 1 | Snare (Head) | 38 |
-| 2 | Snare (Rim) | 37, 40 |
-| 3 | Closed Hi-Hat | 42, 22 |
-| 4 | Open Hi-Hat | 46, 26 |
-| 5 | Low Tom | 43, 58 |
-| 6 | Mid Tom | 47, 45 |
-| 7 | High Tom | 50, 48 |
-| 8 | Ride / Crash | 49, 51, 55, 57, 52 |
+## Loading
 
-## Usage Examples
-
-### Load audio and MIDI
+### Stream Raw Audio + MIDI
 
 ```python
 from datasets import load_dataset
 
-ds = load_dataset("schismaudio/e-gmd")
-
-sample = ds["train"][0]
-print(f"Drummer: {sample['drummer']}, BPM: {sample['bpm']}")
-audio_array = sample["audio"]["array"]
-sr = sample["audio"]["sampling_rate"]
+ds = load_dataset("schismaudio/e-gmd", name="raw", split="train", streaming=True)
+for example in ds:
+    # audio.flac, midi.mid, metadata.json
+    pass
 ```
 
-### Parse MIDI onsets
+### Load Features
 
 ```python
-import io
-import pretty_midi
-
-sample = ds["train"][0]
-midi = pretty_midi.PrettyMIDI(io.BytesIO(sample["midi"]))
-for instrument in midi.instruments:
-    for note in instrument.notes:
-        print(f"Time: {note.start:.3f}s, Pitch: {note.pitch}, Velocity: {note.velocity}")
+ds = load_dataset("schismaudio/e-gmd", name="features", split="train")
+for example in ds:
+    # mel_spectrogram.npy (128, T), onset_targets.npy (26, T), velocity_targets.npy (26, T), params.json
+    pass
 ```
 
-## Dataset Creation
-
-### Source Data
-
-E-GMD builds on the original [Groove MIDI Dataset](https://huggingface.co/datasets/schismaudio/groove-midi-dataset) recorded at Google Magenta. The original performances were captured on a Roland TD-11, with MIDI data preserving exact onset times, pitches, and velocities. For E-GMD, each performance was re-rendered through 43 different drum kits on a **Roland TD-17**, multiplying the timbral diversity of the dataset by 43x.
-
-### Annotations
-
-Annotations are the original MIDI data from the electronic drum kit — no manual annotation was needed. The re-rendering process preserved all timing and velocity information from the original performances.
-
-## Known Limitations
-
-- **Synthesized audio only:** All audio is the electronic drum kit's internal sound engine output, not acoustic drums.
-- **Shared performances:** The underlying MIDI performances are shared across all 43 kit renderings. Models should be evaluated on held-out drummers, not just held-out kit renderings.
-- **Electronic pad dynamics:** Velocity response differs from acoustic drums — pad triggers have fixed dynamic curves.
-- **Large dataset size:** At 444 hours, the full dataset requires significant storage. Consider using streaming mode for exploration.
-
-## Related Datasets
-
-This dataset is part of the [Drum Audio Datasets](https://huggingface.co/collections/schismaudio/drum-audio-datasets) collection by [schismaudio](https://huggingface.co/schismaudio). Related datasets:
-
-- [schismaudio/groove-midi-dataset](https://huggingface.co/datasets/schismaudio/groove-midi-dataset) — The original 13.6-hour GMD with a single kit
-- [schismaudio/stemgmd](https://huggingface.co/datasets/schismaudio/stemgmd) — 1,224 hours of isolated per-instrument stems from 10 acoustic kits
+## Feature Parameters
+
+| Parameter | Value |
+|---|---|
+| Sample rate | 16000 Hz |
+| Hop length | 256 |
+| n_mels | 128 |
+| n_classes | 26 (GM drum pitches 35-60) |
+| FPS | 62.5 |
+
+## 26-Class Drum Mapping
+
+| Class | MIDI | Instrument |
+|---|---|---|
+| 0 | 35 | Acoustic Bass Drum |
+| 1 | 36 | Bass Drum 1 |
+| 2 | 37 | Side Stick |
+| 3 | 38 | Acoustic Snare |
+| 4 | 39 | Hand Clap |
+| 5 | 40 | Electric Snare |
+| 6 | 41 | Low Floor Tom |
+| 7 | 42 | Closed Hi-Hat |
+| 8 | 43 | High Floor Tom |
+| 9 | 44 | Pedal Hi-Hat |
+| 10 | 45 | Low Tom |
+| 11 | 46 | Open Hi-Hat |
+| 12 | 47 | Low-Mid Tom |
+| 13 | 48 | Hi-Mid Tom |
+| 14 | 49 | Crash Cymbal 1 |
+| 15 | 50 | High Tom |
+| 16 | 51 | Ride Cymbal 1 |
+| 17 | 52 | Chinese Cymbal |
+| 18 | 53 | Ride Bell |
+| 19 | 54 | Tambourine |
+| 20 | 55 | Splash Cymbal |
+| 21 | 56 | Cowbell |
+| 22 | 57 | Crash Cymbal 2 |
+| 23 | 58 | Vibraslap |
+| 24 | 59 | Ride Cymbal 2 |
+| 25 | 60 | Hi Bongo |
+
+**Pitch aliases:** MIDI 22 → class 7 (Closed Hi-Hat), MIDI 26 → class 11 (Open Hi-Hat)
 
 ## Citation
 
@@ -181,6 +130,4 @@ This dataset is part of the [Drum Audio Datasets](https://huggingface.co/collect
 
 ## License
 
-This dataset is released under the [Creative Commons Attribution 4.0 International License (CC-BY 4.0)](https://creativecommons.org/licenses/by/4.0/).
-
-You are free to share and adapt this dataset for any purpose, including commercial use, as long as you give appropriate credit to the original authors (Google Magenta).
+CC-BY-4.0