# Music Modeling on Piano-e-Competition (val)

Source: [Music Modeling on Piano-e-Competition (val)](<https://www.sota2.com/research/sota/music-modeling-on-piano-e-competition-val>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:08:05.459Z

Evaluation of Music Modeling performance on the validation split of the Piano-e-Competition dataset, typically measured by Validation NLL.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Music Modeling |
| Dataset (including any stated split) | Piano-e-Competition (val) |
| Metric | Val NLL |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 6 of 6 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Val NLL](<https://www.sota2.com/research/sota/music-modeling-on-piano-e-competition-val.md?metric=Val%20NLL>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Val NLL | Result date | Report paper |
| --- | --- | --- | --- |
| TF with relative global attention | 1.835 | Not recorded | 68fa65341e209131cea08ebc |
| TF with relative local attention | 1.84 | Not recorded | 68fa65341e209131cea08ebc |
| Transformer (TF) baseline | 1.861 | Not recorded | 68fa65341e209131cea08ebc |
| TF with local attention | 1.863 | Not recorded | 68fa65341e209131cea08ebc |
| LSTM with attention | 1.959 | Not recorded | 68fa65341e209131cea08ebc |
| PERFORMANCE RNN (LSTM) | 1.969 | Not recorded | 68fa65341e209131cea08ebc |

### TF with relative global attention

Val NLL: 1.835. Result ID: 698ca1cd0b614d12cff2158c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Implementation](<https://storage.googleapis.com/music-transformer/index.html>)

Recorded evaluation configuration: {"Architecture":"Music Transformer (Relative Global)","Number of layers (L)":"6","Filter size (fs)":"2048","Maximum relative distance (r)":"1024","Sequence length":"2048"}. This record does not assert independent reproduction.

### TF with relative local attention

Val NLL: 1.84. Result ID: 698ca1cd0b614d12cff2158d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Implementation](<https://storage.googleapis.com/music-transformer/index.html>)

Recorded evaluation configuration: {"Architecture":"Music Transformer (Relative Local)","Number of layers (L)":"6","Filter size (fs)":"1024","Maximum relative distance (r)":"2048","Block size (bs)":"512","Sequence length":"2048"}. This record does not assert independent reproduction.

### Transformer (TF) baseline

Val NLL: 1.861. Result ID: 698ca1cd0b614d12cff2158a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa703d1e209131cea49db1.md>)

[Implementation](<https://github.com/tensorflow/tensor2tensor>)

Recorded evaluation configuration: {"Architecture":"Transformer","Number of layers (L)":"6","Hidden size (hs)":"256","Attention size (att)":"512","Filter size (fs)":"2048","Maximum relative distance (r)":"1024","Number of heads (h)":"8","Sequence length":"2048"}. This record does not assert independent reproduction.

### TF with local attention

Val NLL: 1.863. Result ID: 698ca1cd0b614d12cff2158b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64081e209131cea02222.md>)

Recorded evaluation configuration: {"Architecture":"Transformer","Number of layers (L)":"8","Filter size (fs)":"1024","Block size (bs)":"512","Sequence length":"2048"}. This record does not assert independent reproduction.

### LSTM with attention

Val NLL: 1.959. Result ID: 698ca1cd0b614d12cff21589.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

Recorded evaluation configuration: {"Architecture":"LSTM","Number of layers (L)":"3","Hidden size (hs)":"1024","Attention size (att)":"1024","Sequence length":"2048"}. This record does not assert independent reproduction.

### PERFORMANCE RNN (LSTM)

Val NLL: 1.969. Result ID: 698ca1cd0b614d12cff21588.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65341e209131cea08ebc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64be1e209131cea0637b.md>)

Recorded evaluation configuration: {"Architecture":"LSTM","Number of layers (L)":"3","Hidden size (hs)":"1024","Sequence length":"2048"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.