# KG-to-text generation on AGENDA (test) Source: [KG-to-text generation on AGENDA (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:38:21.203Z Evaluation of Knowledge Graph (KG) to text generation performance on the test split of the AGENDA dataset, measured using metrics such as BLEU, METEOR, and CHRF++. 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 | KG-to-text generation | | Dataset (including any stated split) | AGENDA (test) | | Metric | CHRF++ | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 22 of 22 | | 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 - [CHRF++](): Higher is better; unit: PERCENT. - [BLEU](): Higher is better; unit: PERCENT. - [METEOR](): Higher is better; unit: PERCENT. - [BS (F1)](): Higher is better; unit: PERCENT. - [MS](): Higher is better; unit: SCALAR. ## Recorded results | Method | CHRF++ | Result date | Report paper | | --- | --- | --- | --- | | BART-large + STA | 51.63 | Not recorded | 68fa6add1e209131cea2ae83 | | BART-large + LMA | 51.33 | Not recorded | 68fa6add1e209131cea2ae83 | | BART-large | 50.44 | Not recorded | 68fa6add1e209131cea2ae83 | | T5-large + STA | 50.27 | Not recorded | 68fa6add1e209131cea2ae83 | | T5-large + LMA | 49.37 | Not recorded | 68fa6add1e209131cea2ae83 | | T5-base | 48.14 | Not recorded | 68fa6add1e209131cea2ae83 | | T5-large | 48.14 | Not recorded | 68fa6add1e209131cea2ae83 | | BART-base | 48.02 | Not recorded | 68fa6add1e209131cea2ae83 | | CGE-LW | 46.69 | Not recorded | 68fa67cd1e209131cea188b3 | | CGE-LW | 46.69 | Not recorded | 68fa69d51e209131cea24769 | | CGE | 46.47 | Not recorded | 68fa67cd1e209131cea188b3 | | PGE | 46.41 | Not recorded | 68fa67cd1e209131cea188b3 | | Ribeiro et al. | 46.37 | Not recorded | 68fa6add1e209131cea2ae83 | | PGE-LW | 45.79 | Not recorded | 68fa67cd1e209131cea188b3 | | Graformer | 45.43 | Not recorded | 68fa69d51e209131cea24769 | | T5-small | 44.91 | Not recorded | 68fa6add1e209131cea2ae83 | | Local Encoder | 44.7 | Not recorded | 68fa67cd1e209131cea188b3 | | Schmitt et al. | 44.53 | Not recorded | 68fa6add1e209131cea2ae83 | | Global Encoder | 43.95 | Not recorded | 68fa67cd1e209131cea188b3 | ### BART-large + STA CHRF++: 51.63. Result ID: 698e4010ffc3901de09c034a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"BART","Scale":"large","Task-adaptive pretraining":"Supervised Task Adaptation (STA)"}. This record does not assert independent reproduction. ### BART-large + LMA CHRF++: 51.33. Result ID: 698e4010ffc3901de09c0348. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"BART","Scale":"large","Task-adaptive pretraining":"Language Model Adaptation (LMA)"}. This record does not assert independent reproduction. ### BART-large CHRF++: 50.44. Result ID: 698e4010ffc3901de09c0344. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"BART","Scale":"large","Task-adaptive pretraining":"None"}. This record does not assert independent reproduction. ### T5-large + STA CHRF++: 50.27. Result ID: 698e4010ffc3901de09c034b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"T5","Scale":"large","Task-adaptive pretraining":"Supervised Task Adaptation (STA)"}. This record does not assert independent reproduction. ### T5-large + LMA CHRF++: 49.37. Result ID: 698e4010ffc3901de09c0349. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"T5","Scale":"large","Task-adaptive pretraining":"Language Model Adaptation (LMA)"}. This record does not assert independent reproduction. ### T5-base CHRF++: 48.14. Result ID: 698e4010ffc3901de09c0346. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"T5","Scale":"base","Task-adaptive pretraining":"None"}. This record does not assert independent reproduction. ### T5-large CHRF++: 48.14. Result ID: 698e4010ffc3901de09c0347. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"T5","Scale":"large","Task-adaptive pretraining":"None"}. This record does not assert independent reproduction. ### BART-base CHRF++: 48.02. Result ID: 698e4010ffc3901de09c0343. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"BART","Scale":"base","Task-adaptive pretraining":"None"}. This record does not assert independent reproduction. ### CGE-LW CHRF++: 46.69. Result ID: 698d466f8b517eb433204bf9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#L":"6","#H":"8, 8","#P":"69.8M"}. This record does not assert independent reproduction. ### CGE-LW CHRF++: 46.69. Result ID: 698debe5ffc3901de0999359. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Number of parameters":"69.8M"}. This record does not assert independent reproduction. ### CGE CHRF++: 46.47. Result ID: 698d466f8b517eb433204bf7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#L":"6, 3","#H":"8, 8","#P":"61.5M"}. This record does not assert independent reproduction. ### PGE CHRF++: 46.41. Result ID: 698d466f8b517eb433204bf6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#L":"6, 3","#H":"8, 8","#P":"56.1M"}. This record does not assert independent reproduction. ### Ribeiro et al. CHRF++: 46.37. Result ID: 698e4010ffc3901de09c0342. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### PGE-LW CHRF++: 45.79. Result ID: 698d466f8b517eb433204bf8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#L":"6","#H":"8, 8","#P":"69.0M"}. This record does not assert independent reproduction. ### Graformer CHRF++: 45.43. Result ID: 698debe5ffc3901de0999356. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Number of parameters":"36.3M"}. This record does not assert independent reproduction. ### T5-small CHRF++: 44.91. Result ID: 698e4010ffc3901de09c0345. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"T5","Scale":"small","Task-adaptive pretraining":"None"}. This record does not assert independent reproduction. ### Local Encoder CHRF++: 44.7. Result ID: 698d466f8b517eb433204bf5. [Reporting paper]() Recorded evaluation configuration: {"#L":"3","#H":"8","#P":"54.0M"}. This record does not assert independent reproduction. ### Schmitt et al. CHRF++: 44.53. Result ID: 698e4010ffc3901de09c0341. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Global Encoder CHRF++: 43.95. Result ID: 698d466f8b517eb433204bf4. [Reporting paper]() Recorded evaluation configuration: {"#L":"6","#H":"8","#P":"54.4M"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.