# English-to-X Speech-to-text translation on CoVoST 2 Source: [English-to-X Speech-to-text translation on CoVoST 2]() Collection index: [llms.txt]() Data updated: 2026-06-16T10:07:08.519Z Evaluation of Speech-to-text Translation performance on the CoVoST 2 dataset, specifically focusing on translation directions where English is the source language and various other languages (X) are the targets. Metrics reported include COMET and BLEU scores for each target language. 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 | English-to-X Speech-to-text translation | | Dataset (including any stated split) | CoVoST 2 | | Metric | COMET Score (Tr) | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 8 of 8 | | 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 - [COMET Score (Tr)](): Higher is better; unit: SCALAR. - [BLEU Score (Tr)](): Higher is better; unit: SCALAR. - [COMET Score (Ta)](): Higher is better; unit: SCALAR. - [BLEU Score (Ta)](): Higher is better; unit: SCALAR. - [COMET Score (Ja)](): Higher is better; unit: SCALAR. - [BLEU Score (Ja)](): Higher is better; unit: SCALAR. - [COMET Score (Sl)](): Higher is better; unit: SCALAR. - [BLEU Score (Sl)](): Higher is better; unit: SCALAR. - [COMET Score (Lv)](): Higher is better; unit: SCALAR. - [BLEU Score (Lv)](): Higher is better; unit: SCALAR. - [COMET Score (Fa)](): Higher is better; unit: SCALAR. - [BLEU Score (Fa)](): Higher is better; unit: SCALAR. - [COMET Score (Et)](): Higher is better; unit: SCALAR. - [BLEU Score (Et)](): Higher is better; unit: SCALAR. - [COMET Score (De)](): Higher is better; unit: SCALAR. - [BLEU Score (De)](): Higher is better; unit: SCALAR. - [COMET Score (Zh)](): Higher is better; unit: SCALAR. - [BLEU Score (Zh)](): Higher is better; unit: SCALAR. - [COMET Score (Mn)](): Higher is better; unit: SCALAR. - [BLEU Score (Mn)](): Higher is better; unit: SCALAR. - [COMET Score (Sv)](): Higher is better; unit: SCALAR. - [BLEU Score (Sv)](): Higher is better; unit: SCALAR. - [COMET Score (Cy)](): Higher is better; unit: SCALAR. - [BLEU Score (Cy)](): Higher is better; unit: SCALAR. - [COMET Score (Ca)](): Higher is better; unit: SCALAR. - [BLEU Score (Ca)](): Higher is better; unit: SCALAR. - [COMET Score (Ar)](): Higher is better; unit: SCALAR. - [BLEU Score (Ar)](): Higher is better; unit: SCALAR. - [COMET Score (Id)](): Higher is better; unit: SCALAR. - [BLEU Score (Id)](): Higher is better; unit: SCALAR. - [Average COMET Score](): Higher is better; unit: SCALAR. - [Average BLEU Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | COMET Score (Tr) | Result date | Report paper | | --- | --- | --- | --- | | SPRI 2 stage | 78.85 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | SPRI | 78.51 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | Dense FT | 77.28 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | Drop Upcycling | 75.94 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | Noise Upcycling | 75.37 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | LoRA | 74.89 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | BTX | 74.58 | Not recorded | 6a30d8ddd33dfb23e783cf92 | | Naive Upcycling | 73.94 | Not recorded | 6a30d8ddd33dfb23e783cf92 | ### SPRI 2 stage COMET Score (Tr): 78.85. Result ID: 6a312038fbc373592c2e0e16. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. ### SPRI COMET Score (Tr): 78.51. Result ID: 6a312038fbc373592c2e0e15. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. ### Dense FT COMET Score (Tr): 77.28. Result ID: 6a312038fbc373592c2e0e0f. [Reporting paper]() Recorded evaluation configuration: {"Param":"0.7B"}. This record does not assert independent reproduction. ### Drop Upcycling COMET Score (Tr): 75.94. Result ID: 6a312038fbc373592c2e0e12. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. ### Noise Upcycling COMET Score (Tr): 75.37. Result ID: 6a312038fbc373592c2e0e13. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. ### LoRA COMET Score (Tr): 74.89. Result ID: 6a312038fbc373592c2e0e10. [Reporting paper]() Recorded evaluation configuration: {"Param":"0.4B"}. This record does not assert independent reproduction. ### BTX COMET Score (Tr): 74.58. Result ID: 6a312038fbc373592c2e0e14. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. ### Naive Upcycling COMET Score (Tr): 73.94. Result ID: 6a312038fbc373592c2e0e11. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Param":"1.3B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.