QuangDuy/bert-base-stage2-sbert-v2

bert-base-stage2-sbert

SentenceTransformer checkpoint fine-tuned for Vietnamese legal retrieval.

Evaluation

• Truncate dims: [768, 512, 256, 128]

UIT-ViQuAD2.0

method	Accuracy@1	Accuracy@3	Accuracy@5	Accuracy@10	NDCG@3	NDCG@5	NDCG@10	MRR@3	MRR@5	MRR@10	Recall@5	Recall@10	Recall@100	MAP@100
BM25	0.606492	0.75565	0.805643	0.861663	0.694646	0.715313	0.733532	0.673492	0.685004	0.692583	0.805643	0.861663	0.966717	0.69741
768	0.439666	0.633201	0.711546	0.793179	0.552914	0.585221	0.611751	0.525156	0.543106	0.55413	0.711546	0.793179	0.960827	0.561602
512	0.432543	0.621148	0.702096	0.789892	0.543308	0.57661	0.605102	0.516368	0.534824	0.546642	0.702096	0.789892	0.958773	0.554129
256	0.413368	0.605533	0.686207	0.774825	0.525913	0.559013	0.587817	0.498379	0.516671	0.528646	0.686207	0.774825	0.954664	0.536549
128	0.391316	0.572935	0.652376	0.749349	0.497692	0.530387	0.561732	0.471671	0.489798	0.502725	0.652376	0.749349	0.945624	0.511337

Zalo-Legal

method	Accuracy@1	Accuracy@3	Accuracy@5	Accuracy@10	NDCG@3	NDCG@5	NDCG@10	MRR@3	MRR@5	MRR@10	Recall@5	Recall@10	Recall@100	MAP@100
BM25	0.379442	0.616751	0.69797	0.77665	0.515637	0.549114	0.57497	0.48181	0.500402	0.511188	0.696701	0.775381	0.923858	0.517283
768	0.341371	0.524112	0.611675	0.722081	0.447023	0.483089	0.518636	0.421531	0.441201	0.455775	0.611041	0.721447	0.946066	0.46543
512	0.333756	0.506345	0.582487	0.72335	0.43408	0.465382	0.511369	0.409898	0.427475	0.446424	0.581218	0.722716	0.944797	0.456143
256	0.319797	0.506345	0.583756	0.72335	0.430026	0.461954	0.507023	0.404188	0.421827	0.440389	0.583122	0.722716	0.92703	0.449293
128	0.294416	0.48731	0.585025	0.727157	0.407384	0.447667	0.492874	0.380499	0.402771	0.420987	0.584391	0.726523	0.918147	0.428892