# Visual Question Answering on Flickr30k RoQA (test) Source: [Visual Question Answering on Flickr30k RoQA (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:36:43.274Z Evaluation of model performance on the test split of the Flickr30k RoQA dataset for visual question answering, reporting generation metrics such as BERTScore, ROUGE, and BLEU. 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 | Visual Question Answering | | Dataset (including any stated split) | Flickr30k RoQA (test) | | Metric | BERTScore P | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [BERTScore P](): Higher is better; unit: PERCENT. - [BERTScore R](): Higher is better; unit: PERCENT. - [BERTScore F1](): Higher is better; unit: PERCENT. - [ROUGE-L P](): Higher is better; unit: PERCENT. - [ROUGE-L R](): Higher is better; unit: PERCENT. - [ROUGE-L F1](): Higher is better; unit: PERCENT. - [BLEU Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | BERTScore P | Result date | Report paper | | --- | --- | --- | --- | | Qwen2-VL-RoVQA | 80.43 | Not recorded | 69438a63423ea4226572927d | | LLaMA-3.2-RoVQA | 75.97 | Not recorded | 69438a63423ea4226572927d | | Qwen2-VL | 71.59 | Not recorded | 69438a63423ea4226572927d | | LLaMA-3.2 | 61 | Not recorded | 69438a63423ea4226572927d | | LLaVA-v1.6 | 58.17 | Not recorded | 69438a63423ea4226572927d | | LLaVA-v1.6-RoVQA | 49.5 | Not recorded | 69438a63423ea4226572927d | ### Qwen2-VL-RoVQA BERTScore P: 80.43. Result ID: 6996bd665698665ec4085263. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Adaptation Strategy":"RoVQA Adapter"}. This record does not assert independent reproduction. ### LLaMA-3.2-RoVQA BERTScore P: 75.97. Result ID: 6996bd665698665ec408525f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"11B","Adaptation Strategy":"RoVQA Adapter"}. This record does not assert independent reproduction. ### Qwen2-VL BERTScore P: 71.59. Result ID: 6996bd665698665ec4085262. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"7B","Adaptation Strategy":"None (Base Model)"}. This record does not assert independent reproduction. ### LLaMA-3.2 BERTScore P: 61. Result ID: 6996bd665698665ec408525e. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"11B","Adaptation Strategy":"None (Base Model)"}. This record does not assert independent reproduction. ### LLaVA-v1.6 BERTScore P: 58.17. Result ID: 6996bd665698665ec4085260. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"7B","Adaptation Strategy":"None (Base Model)"}. This record does not assert independent reproduction. ### LLaVA-v1.6-RoVQA BERTScore P: 49.5. Result ID: 6996bd665698665ec4085261. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Adaptation Strategy":"RoVQA Adapter"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.