# End-to-End Speech Synthesis on End-to-End Speech Synthesis Tacotron2 pipeline

Source: [End-to-End Speech Synthesis on End-to-End Speech Synthesis Tacotron2 pipeline](<https://www.sota2.com/research/sota/end-to-end-speech-synthesis-on-end-to-end-speech-synthesis-tacotron2-pipeline>)

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

Data updated: 2026-02-26T05:50:04.424Z

Evaluation of End-to-End Speech Synthesis performance using data derived from the Tacotron2 pipeline, measured primarily using Mean Opinion Score (MOS).

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 | End-to-End Speech Synthesis |
| Dataset (including any stated split) | End-to-End Speech Synthesis Tacotron2 pipeline |
| Metric | MOS |
| Unit | SCALAR |
| Direction | Higher 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

- [MOS](<https://www.sota2.com/research/sota/end-to-end-speech-synthesis-on-end-to-end-speech-synthesis-tacotron2-pipeline.md?metric=MOS>): Higher is better; unit: SCALAR.

## Recorded results

| Method | MOS | Result date | Report paper |
| --- | --- | --- | --- |
| Ground Truth | 4.23 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V1 | 4.18 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V2 | 4.12 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V3 | 4.02 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V1 | 3.91 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V3 | 3.89 | Not recorded | 68fa68a01e209131cea1d3a7 |
| HiFi-GAN V2 | 3.88 | Not recorded | 68fa68a01e209131cea1d3a7 |
| WaveGlow | 3.69 | Not recorded | 68fa68a01e209131cea1d3a7 |
| WaveGlow | 3.66 | Not recorded | 68fa68a01e209131cea1d3a7 |

### Ground Truth

MOS: 4.23. Result ID: 698d7a718b517eb43321d16a.

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

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

### HiFi-GAN V1

MOS: 4.18. Result ID: 698d7a718b517eb43321d170.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":true}. This record does not assert independent reproduction.

### HiFi-GAN V2

MOS: 4.12. Result ID: 698d7a718b517eb43321d171.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":true}. This record does not assert independent reproduction.

### HiFi-GAN V3

MOS: 4.02. Result ID: 698d7a718b517eb43321d172.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":true}. This record does not assert independent reproduction.

### HiFi-GAN V1

MOS: 3.91. Result ID: 698d7a718b517eb43321d16c.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":false}. This record does not assert independent reproduction.

### HiFi-GAN V3

MOS: 3.89. Result ID: 698d7a718b517eb43321d16e.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":false}. This record does not assert independent reproduction.

### HiFi-GAN V2

MOS: 3.88. Result ID: 698d7a718b517eb43321d16d.

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

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

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

Recorded evaluation configuration: {"Fine-tuning":false}. This record does not assert independent reproduction.

### WaveGlow

MOS: 3.69. Result ID: 698d7a718b517eb43321d16b.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa65021e209131cea07da6.md>)

[Implementation](<https://nv-adlr.github.io/WaveGlow>)

Recorded evaluation configuration: {"Fine-tuning":false}. This record does not assert independent reproduction.

### WaveGlow

MOS: 3.66. Result ID: 698d7a718b517eb43321d16f.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa65021e209131cea07da6.md>)

[Implementation](<https://nv-adlr.github.io/WaveGlow>)

Recorded evaluation configuration: {"Fine-tuning":true}. This record does not assert independent reproduction.

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