# Neural Vocoding on LibriTTS-R clean-100 (test)

Source: [Neural Vocoding on LibriTTS-R clean-100 (test)](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test>)

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

Data updated: 2026-05-26T11:23:34.332Z

Evaluation of Neural Vocoding performance on the test split of the LibriTTS-R clean-100 subset.

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 | Neural Vocoding |
| Dataset (including any stated split) | LibriTTS-R clean-100 (test) |
| Metric | NISQA |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 20 of 20 |
| 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

- [NISQA](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=NISQA>): Higher is better; unit: SCALAR.

- [Training Time (hours)](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=Training%20Time%20(hours)>): Lower is better; unit: TIME.

- [RTF (GPU)](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=RTF%20(GPU)>): Lower is better; unit: SCALAR.

- [RTF (CPU)](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=RTF%20(CPU)>): Lower is better; unit: SCALAR.

- [UTMOS](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=UTMOS>): Higher is better; unit: SCALAR.

- [MCD](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=MCD>): Lower is better; unit: SCALAR.

- [log F0 RMSE](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=log%20F0%20RMSE>): Lower is better; unit: SCALAR.

- [Model Size (M)](<https://www.sota2.com/research/sota/neural-vocoding-on-libritts-r-clean-100-test.md?metric=Model%20Size%20(M)>): Lower is better; unit: SIZE.

## Recorded results

| Method | NISQA | Result date | Report paper |
| --- | --- | --- | --- |
| Ground Truth | 4.08 | Not recorded | 6a152943dbd7150506f7e10e |
| WaveFit | 4.02 | Not recorded | 6a152943dbd7150506f7e10e |
| GAN-WaveNeXt 2 | 4.01 | Not recorded | 6a152943dbd7150506f7e10e |
| GAN-WaveNeXt 2 | 4.01 | Not recorded | 6a152943dbd7150506f7e10e |
| HiFi-GAN V1 | 3.99 | Not recorded | 6a152943dbd7150506f7e10e |
| WaveFit | 3.97 | Not recorded | 6a152943dbd7150506f7e10e |
| GAN-WaveNeXt 2 | 3.92 | Not recorded | 6a152943dbd7150506f7e10e |
| WaveFit | 3.91 | Not recorded | 6a152943dbd7150506f7e10e |
| Diff-WaveNeXt 2 | 3.81 | Not recorded | 6a152943dbd7150506f7e10e |
| WaveFit | 3.8 | Not recorded | 6a152943dbd7150506f7e10e |
| GAN-WaveNeXt 2 | 3.77 | Not recorded | 6a152943dbd7150506f7e10e |
| FastDiff | 3.67 | Not recorded | 6a152943dbd7150506f7e10e |
| Diff-WaveNeXt 2 | 3.45 | Not recorded | 6a152943dbd7150506f7e10e |
| FastDiff | 3.43 | Not recorded | 6a152943dbd7150506f7e10e |
| WaveNeXt | 3.16 | Not recorded | 6a152943dbd7150506f7e10e |

### Ground Truth

NISQA: 4.08. Result ID: 6a1582b2fbc373592c0de3f5.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### WaveFit

NISQA: 4.02. Result ID: 6a1582b2fbc373592c0de3fd.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d901e209131cea3aa80.md>)

Recorded evaluation configuration: {"Iterations":"5"}. This record does not assert independent reproduction.

### GAN-WaveNeXt 2

NISQA: 4.01. Result ID: 6a1582b2fbc373592c0de3fc.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Iterations":"4"}. This record does not assert independent reproduction.

### GAN-WaveNeXt 2

NISQA: 4.01. Result ID: 6a1582b2fbc373592c0de3fe.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Iterations":"5"}. This record does not assert independent reproduction.

### HiFi-GAN V1

NISQA: 3.99. Result ID: 6a1582b2fbc373592c0de3ff.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68a01e209131cea1d3a7.md>)

[Implementation](<https://github.com/jik876/hifi-gan>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### WaveFit

NISQA: 3.97. Result ID: 6a1582b2fbc373592c0de3fb.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d901e209131cea3aa80.md>)

Recorded evaluation configuration: {"Iterations":"4"}. This record does not assert independent reproduction.

### GAN-WaveNeXt 2

NISQA: 3.92. Result ID: 6a1582b2fbc373592c0de3fa.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Iterations":"3"}. This record does not assert independent reproduction.

### WaveFit

NISQA: 3.91. Result ID: 6a1582b2fbc373592c0de3f9.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d901e209131cea3aa80.md>)

Recorded evaluation configuration: {"Iterations":"3"}. This record does not assert independent reproduction.

### Diff-WaveNeXt 2

NISQA: 3.81. Result ID: 6a1582b2fbc373592c0de403.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Sub-modeling strategy":"with"}. This record does not assert independent reproduction.

### WaveFit

NISQA: 3.8. Result ID: 6a1582b2fbc373592c0de3f7.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d901e209131cea3aa80.md>)

Recorded evaluation configuration: {"Iterations":"2"}. This record does not assert independent reproduction.

### GAN-WaveNeXt 2

NISQA: 3.77. Result ID: 6a1582b2fbc373592c0de3f8.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Iterations":"2"}. This record does not assert independent reproduction.

### FastDiff

NISQA: 3.67. Result ID: 6a1582b2fbc373592c0de402.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c551e209131cea335b7.md>)

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

Recorded evaluation configuration: {"Sub-modeling strategy":"with"}. This record does not assert independent reproduction.

### Diff-WaveNeXt 2

NISQA: 3.45. Result ID: 6a1582b2fbc373592c0de401.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Sub-modeling strategy":"without"}. This record does not assert independent reproduction.

### FastDiff

NISQA: 3.43. Result ID: 6a1582b2fbc373592c0de400.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c551e209131cea335b7.md>)

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

Recorded evaluation configuration: {"Sub-modeling strategy":"without"}. This record does not assert independent reproduction.

### WaveNeXt

NISQA: 3.16. Result ID: 6a1582b2fbc373592c0de3f6.

[Reporting paper](<https://www.sota2.com/research/paper/6a152943dbd7150506f7e10e.md>)

Recorded evaluation configuration: {"Iterations":"1"}. This record does not assert independent reproduction.

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