# Music Generation on FMACaps

Source: [Music Generation on FMACaps](<https://www.sota2.com/research/sota/music-generation-on-fmacaps>)

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

Data updated: 2026-02-26T05:44:04.171Z

Evaluation of Music Generation performance using the FMACaps dataset, assessed using metrics like FD, FAD, and KL divergence.

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 Generation |
| Dataset (including any stated split) | FMACaps |
| Metric | FD |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 9 of 9 |
| 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

- [FD](<https://www.sota2.com/research/sota/music-generation-on-fmacaps.md?metric=FD>): Lower is better; unit: SCALAR.

- [FAD](<https://www.sota2.com/research/sota/music-generation-on-fmacaps.md?metric=FAD>): Lower is better; unit: SCALAR.

- [KL Divergence](<https://www.sota2.com/research/sota/music-generation-on-fmacaps.md?metric=KL%20Divergence>): Lower is better; unit: SCALAR.

## Recorded results

| Method | FD | Result date | Report paper |
| --- | --- | --- | --- |
| AudioLDM2 | 19.99 | Not recorded | 68fa805b1e209131cea9cc5d |
| MusicGen-M | 22.61 | Not recorded | 68fa805b1e209131cea9cc5d |
| MusicGen-S | 23.21 | Not recorded | 68fa805b1e209131cea9cc5d |
| Mustango | 24.24 | Not recorded | 68fa805b1e209131cea9cc5d |
| Tango | 24.48 | Not recorded | 68fa805b1e209131cea9cc5d |
| Mustango | 25.18 | Not recorded | 68fa805b1e209131cea9cc5d |
| Tango | 26.31 | Not recorded | 68fa805b1e209131cea9cc5d |
| Tango | 28.32 | Not recorded | 68fa805b1e209131cea9cc5d |
| Tango | 28.81 | Not recorded | 68fa805b1e209131cea9cc5d |

### AudioLDM2

FD: 19.99. Result ID: 698b95bc7a967eddb76fcc03.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73261e209131cea5ab21.md>)

[Implementation](<https://audioldm.github.io/audioldm2>)

Recorded evaluation configuration: {"Datasets":"-","Pre-trained":false,"Parameters":"346M"}. This record does not assert independent reproduction.

### MusicGen-M

FD: 22.61. Result ID: 698b95bc7a967eddb76fcc02.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72031e209131cea5421d.md>)

[Implementation](<https://www.sota2.com/github.com/facebookresearch/audiocraft>)

Recorded evaluation configuration: {"Datasets":"-","Pre-trained":false,"Parameters":"1.5B"}. This record does not assert independent reproduction.

### MusicGen-S

FD: 23.21. Result ID: 698b95bc7a967eddb76fcc01.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72031e209131cea5421d.md>)

[Implementation](<https://www.sota2.com/github.com/facebookresearch/audiocraft>)

Recorded evaluation configuration: {"Datasets":"-","Pre-trained":false,"Parameters":"300M"}. This record does not assert independent reproduction.

### Mustango

FD: 24.24. Result ID: 698b95bc7a967eddb76fcc08.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Implementation](<https://github.com/amaai-lab/mustango>)

Recorded evaluation configuration: {"Datasets":"MusicBench","Pre-trained":false,"Parameters":"1.4B"}. This record does not assert independent reproduction.

### Tango

FD: 24.48. Result ID: 698b95bc7a967eddb76fcc07.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f771e209131cea45c2c.md>)

[Implementation](<https://github.com/declare-lab/tango>)

Recorded evaluation configuration: {"Datasets":"MusicBench","Pre-trained":true,"Parameters":"866M"}. This record does not assert independent reproduction.

### Mustango

FD: 25.18. Result ID: 698b95bc7a967eddb76fcc09.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Implementation](<https://github.com/amaai-lab/mustango>)

Recorded evaluation configuration: {"Datasets":"MusicBench","Pre-trained":true,"Parameters":"1.4B"}. This record does not assert independent reproduction.

### Tango

FD: 26.31. Result ID: 698b95bc7a967eddb76fcc06.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f771e209131cea45c2c.md>)

[Implementation](<https://github.com/declare-lab/tango>)

Recorded evaluation configuration: {"Datasets":"MusicBench","Pre-trained":false,"Parameters":"866M"}. This record does not assert independent reproduction.

### Tango

FD: 28.32. Result ID: 698b95bc7a967eddb76fcc04.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f771e209131cea45c2c.md>)

[Implementation](<https://github.com/declare-lab/tango>)

Recorded evaluation configuration: {"Datasets":"MusicCaps","Pre-trained":false,"Parameters":"866M"}. This record does not assert independent reproduction.

### Tango

FD: 28.81. Result ID: 698b95bc7a967eddb76fcc05.

[Reporting paper](<https://www.sota2.com/research/paper/68fa805b1e209131cea9cc5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f771e209131cea45c2c.md>)

[Implementation](<https://github.com/declare-lab/tango>)

Recorded evaluation configuration: {"Datasets":"MusicCaps","Pre-trained":true,"Parameters":"866M"}. This record does not assert independent reproduction.

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