# Grapheme-to-Phoneme conversion on CMUDict 0.7b (test)

Source: [Grapheme-to-Phoneme conversion on CMUDict 0.7b (test)](<https://www.sota2.com/research/sota/grapheme-to-phoneme-conversion-on-cmudict-0-7b-test>)

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

Data updated: 2026-02-26T05:55:00.372Z

Evaluation of Grapheme-to-Phoneme conversion accuracy on the test split of the CMUDict 0.7b dataset, typically reporting Phoneme Error Rate (PER) and Word Error Rate (WER).

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 | Grapheme-to-Phoneme conversion |
| Dataset (including any stated split) | CMUDict 0.7b (test) |
| Metric | PER |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 4 of 4 |
| 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

- [PER](<https://www.sota2.com/research/sota/grapheme-to-phoneme-conversion-on-cmudict-0-7b-test.md?metric=PER>): Lower is better; unit: PERCENT.

- [WER](<https://www.sota2.com/research/sota/grapheme-to-phoneme-conversion-on-cmudict-0-7b-test.md?metric=WER>): Lower is better; unit: PERCENT.

## Recorded results

| Method | PER | Result date | Report paper |
| --- | --- | --- | --- |
| Token-Level Ensemble Distillation | 4.6 | Not recorded | 68fa66391e209131cea0ef65 |
| CNN with NSGD | 5.58 | Not recorded | 68fa66391e209131cea0ef65 |
| Bi-LSTM + n-gram | 5.76 | Not recorded | 68fa66391e209131cea0ef65 |
| Sequitur G2P | 6.12 | Not recorded | 68fa66391e209131cea0ef65 |

### Token-Level Ensemble Distillation

PER: 4.6. Result ID: 698ce2f28b517eb4331d7270.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66391e209131cea0ef65.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66391e209131cea0ef65.md>)

Recorded evaluation configuration: {"Backbone":"Transformer","Layers":"6-6"}. This record does not assert independent reproduction.

### CNN with NSGD

PER: 5.58. Result ID: 698ce2f28b517eb4331d726f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66391e209131cea0ef65.md>)

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

### Bi-LSTM + n-gram

PER: 5.76. Result ID: 698ce2f28b517eb4331d726e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66391e209131cea0ef65.md>)

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

### Sequitur G2P

PER: 6.12. Result ID: 698ce2f28b517eb4331d726d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66391e209131cea0ef65.md>)

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

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