# Grapheme-to-Phoneme Conversion on Internal dataset Source: [Grapheme-to-Phoneme Conversion on Internal dataset]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:55:00.391Z Evaluation of Grapheme-to-Phoneme (G2P) conversion performance on the Internal dataset, typically measured using 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) | Internal dataset | | Metric | PER | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 2 of 2 | | 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](): Lower is better; unit: SCALAR. - [WER](): Lower is better; unit: SCALAR. ## Recorded results | Method | PER | Result date | Report paper | | --- | --- | --- | --- | | Token-Level Ensemble Distillation | 0.0304 | Not recorded | 68fa66391e209131cea0ef65 | | CNN with NSGD | 0.0379 | Not recorded | 68fa66391e209131cea0ef65 | ### Token-Level Ensemble Distillation PER: 0.0304. Result ID: 698ce2f28b517eb4331d7272. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CNN with NSGD PER: 0.0379. Result ID: 698ce2f28b517eb4331d7271. [Reporting paper]() Recorded evaluation configuration: {"reproduced":"true"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.