NER-bert-base-multilingual-cased

This model is a fine-tuned version of google-bert/bert-base-multilingual-cased on the hts98/UIT dataset. It achieves the following results on the evaluation set:

• Loss: 2.3492
• Precision: 0.6038
• Recall: 0.6459
• F1: 0.6241
• Accuracy: 0.7757

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 120.0

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	487	0.8134	0.4221	0.5292	0.4696	0.7412
1.0691	2.0	974	0.7439	0.4565	0.5744	0.5087	0.7600
0.6796	3.0	1461	0.8021	0.4755	0.5906	0.5268	0.7500
0.5266	4.0	1948	0.8266	0.4883	0.6171	0.5452	0.7569
0.4087	5.0	2435	0.8820	0.5043	0.6238	0.5577	0.7584
0.314	6.0	2922	0.8884	0.5110	0.6241	0.5619	0.7588
0.254	7.0	3409	0.9710	0.5112	0.6261	0.5628	0.7593
0.2096	8.0	3896	1.0743	0.5137	0.6272	0.5648	0.7622
0.1786	9.0	4383	1.1286	0.5182	0.6255	0.5668	0.7571
0.1486	10.0	4870	1.1630	0.5240	0.6306	0.5724	0.7545
0.132	11.0	5357	1.1934	0.5322	0.6278	0.5760	0.7606
0.1098	12.0	5844	1.1862	0.5380	0.6188	0.5756	0.7602
0.094	13.0	6331	1.3724	0.5295	0.6325	0.5764	0.7506
0.084	14.0	6818	1.3746	0.5304	0.6258	0.5742	0.7532
0.0758	15.0	7305	1.3000	0.5157	0.6333	0.5685	0.7581
0.0694	16.0	7792	1.4195	0.5486	0.6306	0.5867	0.7593
0.062	17.0	8279	1.4974	0.5234	0.6300	0.5718	0.7466
0.0543	18.0	8766	1.5014	0.5347	0.6199	0.5742	0.7568
0.0471	19.0	9253	1.5165	0.5373	0.6227	0.5769	0.7546
0.0449	20.0	9740	1.5719	0.5277	0.6278	0.5734	0.7568
0.0451	21.0	10227	1.5307	0.5582	0.6297	0.5918	0.7607
0.039	22.0	10714	1.5783	0.5437	0.6317	0.5844	0.7572
0.0363	23.0	11201	1.6342	0.5376	0.6303	0.5803	0.7542
0.0326	24.0	11688	1.6417	0.5590	0.6272	0.5911	0.7597
0.0296	25.0	12175	1.6685	0.5414	0.6389	0.5861	0.7587
0.0283	26.0	12662	1.7347	0.5571	0.6331	0.5927	0.7602
0.0277	27.0	13149	1.6560	0.5675	0.6423	0.6026	0.7632
0.025	28.0	13636	1.7497	0.5722	0.6361	0.6025	0.7614
0.0241	29.0	14123	1.7110	0.5652	0.6367	0.5988	0.7638
0.0242	30.0	14610	1.7947	0.5642	0.6297	0.5951	0.7647
0.0219	31.0	15097	1.8283	0.5607	0.6283	0.5926	0.7565
0.0193	32.0	15584	1.8161	0.5690	0.6278	0.5969	0.7648
0.0185	33.0	16071	1.8462	0.5564	0.6347	0.5930	0.7609
0.0195	34.0	16558	1.9018	0.5508	0.6280	0.5869	0.7558
0.0181	35.0	17045	1.8523	0.5638	0.6356	0.5975	0.7597
0.0182	36.0	17532	1.8344	0.5770	0.6328	0.6036	0.7611
0.0153	37.0	18019	1.8465	0.5760	0.6331	0.6032	0.7669
0.0142	38.0	18506	1.8911	0.5679	0.6238	0.5945	0.7632
0.0142	39.0	18993	1.8849	0.5790	0.6241	0.6007	0.7623
0.0151	40.0	19480	1.8399	0.5722	0.6255	0.5977	0.7665
0.0148	41.0	19967	1.8430	0.5782	0.6163	0.5966	0.7649
0.0138	42.0	20454	1.8764	0.5544	0.6278	0.5888	0.7691
0.0147	43.0	20941	1.9270	0.5717	0.6345	0.6015	0.7666
0.0148	44.0	21428	1.8888	0.5621	0.6227	0.5909	0.7711
0.0123	45.0	21915	1.8993	0.5552	0.6225	0.5869	0.7653
0.0115	46.0	22402	1.9475	0.5647	0.6353	0.5979	0.7645
0.0107	47.0	22889	1.9949	0.5778	0.6359	0.6054	0.7674
0.0098	48.0	23376	1.9607	0.5704	0.6275	0.5976	0.7681
0.012	49.0	23863	1.9185	0.5793	0.6518	0.6134	0.7676
0.0117	50.0	24350	1.9814	0.5729	0.6409	0.6050	0.7698
0.0093	51.0	24837	2.0354	0.5761	0.6409	0.6067	0.7662
0.0082	52.0	25324	1.9876	0.5937	0.6442	0.6179	0.7683
0.0077	53.0	25811	2.0616	0.6078	0.6345	0.6208	0.7691
0.0087	54.0	26298	1.9790	0.5634	0.6367	0.5978	0.7653
0.0102	55.0	26785	2.0688	0.5754	0.6392	0.6056	0.7678
0.0073	56.0	27272	1.9601	0.5863	0.6300	0.6073	0.7679
0.0087	57.0	27759	2.0415	0.5791	0.6412	0.6085	0.7683
0.0082	58.0	28246	2.0774	0.5687	0.6395	0.6020	0.7666
0.0056	59.0	28733	2.0773	0.5822	0.6322	0.6062	0.7637
0.0076	60.0	29220	2.1045	0.5968	0.6386	0.6170	0.7695
0.0071	61.0	29707	2.0994	0.5922	0.6278	0.6095	0.7682
0.0076	62.0	30194	2.0937	0.5795	0.6426	0.6094	0.7650
0.0082	63.0	30681	2.0307	0.5775	0.6381	0.6063	0.7683
0.0068	64.0	31168	2.1657	0.5820	0.6353	0.6075	0.7597
0.0065	65.0	31655	2.0142	0.5850	0.6448	0.6134	0.7692
0.0062	66.0	32142	2.1379	0.5777	0.6381	0.6064	0.7602
0.0059	67.0	32629	2.1319	0.5837	0.6426	0.6117	0.7631
0.0053	68.0	33116	2.1246	0.5761	0.6361	0.6046	0.7682
0.0049	69.0	33603	2.1514	0.5807	0.6381	0.6080	0.7657
0.0037	70.0	34090	2.1636	0.5839	0.6400	0.6107	0.7680
0.0053	71.0	34577	2.1478	0.5853	0.6266	0.6053	0.7639
0.0051	72.0	35064	2.1522	0.5779	0.6403	0.6075	0.7688
0.0047	73.0	35551	2.1609	0.5831	0.6381	0.6093	0.7671
0.0036	74.0	36038	2.1757	0.6001	0.6414	0.6201	0.7706
0.004	75.0	36525	2.2280	0.5909	0.6445	0.6165	0.7662
0.0036	76.0	37012	2.2199	0.6016	0.6375	0.6190	0.7710
0.0036	77.0	37499	2.1810	0.5852	0.6409	0.6118	0.7685
0.0043	78.0	37986	2.2161	0.5848	0.6364	0.6095	0.7689
0.0039	79.0	38473	2.1878	0.5748	0.6467	0.6087	0.7694
0.0052	80.0	38960	2.2712	0.5874	0.6308	0.6083	0.7653
0.0034	81.0	39447	2.2645	0.5893	0.6386	0.6130	0.7658
0.0027	82.0	39934	2.2353	0.5995	0.6336	0.6161	0.7651
0.0026	83.0	40421	2.3131	0.5851	0.6356	0.6093	0.7630
0.0017	84.0	40908	2.2798	0.5800	0.6437	0.6102	0.7660
0.0022	85.0	41395	2.3181	0.5879	0.6395	0.6126	0.7637
0.0032	86.0	41882	2.2964	0.5986	0.6364	0.6169	0.7696
0.003	87.0	42369	2.2509	0.5993	0.6420	0.6199	0.7665
0.003	88.0	42856	2.2512	0.6042	0.6386	0.6210	0.7705
0.0027	89.0	43343	2.2787	0.5812	0.6467	0.6122	0.7695
0.0016	90.0	43830	2.2573	0.5861	0.6426	0.6130	0.7653
0.0028	91.0	44317	2.2477	0.5963	0.6467	0.6205	0.7694
0.0022	92.0	44804	2.2446	0.5865	0.6493	0.6163	0.7652
0.0017	93.0	45291	2.2529	0.5917	0.6462	0.6177	0.7661
0.0017	94.0	45778	2.2624	0.5933	0.6400	0.6158	0.7650
0.0015	95.0	46265	2.2784	0.5969	0.6364	0.6160	0.7650
0.0012	96.0	46752	2.3038	0.5859	0.6456	0.6143	0.7629
0.0019	97.0	47239	2.3129	0.5861	0.6501	0.6164	0.7649
0.001	98.0	47726	2.3077	0.5912	0.6420	0.6155	0.7682
0.0009	99.0	48213	2.3493	0.5907	0.6440	0.6162	0.7633
0.0015	100.0	48700	2.3195	0.6003	0.6437	0.6212	0.7701
0.001	101.0	49187	2.3444	0.5956	0.6495	0.6214	0.7711
0.0008	102.0	49674	2.4047	0.5915	0.6417	0.6156	0.7639
0.0011	103.0	50161	2.3442	0.5796	0.6434	0.6098	0.7672
0.0009	104.0	50648	2.3378	0.5919	0.6423	0.6160	0.7682
0.0011	105.0	51135	2.3191	0.6018	0.6431	0.6218	0.7703
0.0007	106.0	51622	2.3766	0.5896	0.6451	0.6161	0.7683
0.0004	107.0	52109	2.3492	0.6038	0.6459	0.6241	0.7757
0.0008	108.0	52596	2.3653	0.5975	0.6462	0.6209	0.7681
0.0005	109.0	53083	2.3852	0.5992	0.6437	0.6206	0.7692
0.0005	110.0	53570	2.4063	0.6053	0.6406	0.6224	0.7685
0.0008	111.0	54057	2.4257	0.6007	0.6395	0.6195	0.7683
0.0009	112.0	54544	2.4032	0.5993	0.6437	0.6207	0.7700
0.0006	113.0	55031	2.3878	0.5967	0.6442	0.6196	0.7707
0.0003	114.0	55518	2.3939	0.6013	0.6423	0.6211	0.7713
0.0003	115.0	56005	2.4125	0.5980	0.6400	0.6183	0.7703
0.0003	116.0	56492	2.4203	0.5957	0.6456	0.6197	0.7706
0.0003	117.0	56979	2.4104	0.6	0.6426	0.6206	0.7707
0.0004	118.0	57466	2.4210	0.6004	0.6445	0.6217	0.7696
0.0004	119.0	57953	2.4213	0.5990	0.6428	0.6202	0.7692
0.0004	120.0	58440	2.4216	0.5993	0.6423	0.6200	0.7694

Framework versions

• Transformers 4.32.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 3.1.0
• Tokenizers 0.13.3