Legato 2
Full-page optical music recognition. A detector finds the systems on a page and a transcriber reads each one in turn, conditioned on the ABC it has already produced for the systems above it, so the notation stays consistent down the page.
Both halves live here: model.safetensors is the transcriber, segmenter.pt the YOLO line
detector. They are versioned together because a change to either one changes the result.
Usage
from legato.pipeline import Legato2
omr = Legato2.from_pretrained("guangyangmusic/legato-2", device="cuda:0")
abc = omr("page.png")
Decoding follows the paper by default: greedy (beam size 1), repetition penalty 1.1, maximum
length 2048; segmentation uses confidence 0.7 at image size 640. All of these can be
overridden per call, for example omr("page.png", beam_size=10, yolo_conf=0.5).
A PDF can be transcribed from the command line, where it is first rendered to one tall image so a multi-page score is read as a single piece:
python scripts/run_legato2.py --image_path score.pdf --model_path guangyangmusic/legato-2
Checkpoint
legato-segment-bytes-tokenizer-lr-3e-4-cosine-e5-b32-v4096: byte-level tokenizer with a
4096-entry vocabulary, learning rate 3e-4 on a cosine schedule, five epochs at batch size 32.
Evaluation
OMR-NED, lower is better. Scored with musicdiff 5.2 on music21 9.7.1; the ABC is converted to MusicXML with abc2xml followed by a MuseScore 3.6.2 rewrite, which is part of the measurement and shifts the figures by one to two points if skipped.
| Dataset | Legato 2 | Legato 1 | Audiveris | Gemini 3.1 Pro |
|---|---|---|---|---|
| Camera String Quartets | 31.6 | 58.3 | 69.3 | 94.8 |
| Rendered String Quartets | 17.1 | 32.6 | 36.8 | 93.5 |
| Camera OpenScore Lieder | 38.3 | 49.0 | 52.1 | 87.2 |
| Rendered OpenScore Lieder | 27.6 | 34.4 | 29.9 | 86.7 |
| IMSLP Piano Scores | 36.6 | 43.8 | 66.0 | 89.1 |
| PDMX-Synth test | 21.7 | 27.4 | 34.7 | 85.2 |
- Downloads last month
- 6