scenario-non-kd-pre-ner-full-mdeberta_data-univner_full44

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1878
• Precision: 0.7925
• Recall: 0.8189
• F1: 0.8055
• Accuracy: 0.9794

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: 32
• eval_batch_size: 32
• seed: 44
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2646	0.2910	500	0.1977	0.3184	0.2992	0.3085	0.9376
0.1629	0.5821	1000	0.1272	0.5005	0.6191	0.5535	0.9577
0.1027	0.8731	1500	0.1002	0.6168	0.7074	0.6590	0.9672
0.0757	1.1641	2000	0.1021	0.6325	0.7381	0.6813	0.9688
0.0645	1.4552	2500	0.0843	0.7217	0.7640	0.7422	0.9748
0.0593	1.7462	3000	0.0784	0.7325	0.7780	0.7545	0.9754
0.0521	2.0373	3500	0.0892	0.7262	0.7931	0.7582	0.9754
0.0411	2.3283	4000	0.0822	0.7397	0.8100	0.7732	0.9768
0.0378	2.6193	4500	0.0854	0.7364	0.8207	0.7763	0.9772
0.0366	2.9104	5000	0.0826	0.7439	0.8160	0.7783	0.9771
0.0284	3.2014	5500	0.0916	0.7535	0.7964	0.7744	0.9772
0.0254	3.4924	6000	0.0912	0.7575	0.7992	0.7778	0.9776
0.0253	3.7835	6500	0.0901	0.7733	0.8107	0.7916	0.9787
0.0236	4.0745	7000	0.1009	0.7784	0.7986	0.7883	0.9784
0.0187	4.3655	7500	0.1032	0.7917	0.8045	0.7981	0.9789
0.019	4.6566	8000	0.1002	0.7507	0.8211	0.7843	0.9776
0.0179	4.9476	8500	0.1002	0.7706	0.8084	0.7890	0.9784
0.0137	5.2386	9000	0.1091	0.7527	0.8173	0.7837	0.9773
0.0134	5.5297	9500	0.1038	0.7727	0.8160	0.7938	0.9783
0.014	5.8207	10000	0.1148	0.7595	0.8120	0.7849	0.9778
0.0129	6.1118	10500	0.1148	0.7657	0.8198	0.7918	0.9780
0.0106	6.4028	11000	0.1178	0.7671	0.8185	0.7920	0.9779
0.0099	6.6938	11500	0.1092	0.7758	0.8159	0.7954	0.9785
0.0102	6.9849	12000	0.1170	0.7681	0.8156	0.7911	0.9775
0.0079	7.2759	12500	0.1235	0.7589	0.8280	0.7920	0.9775
0.0083	7.5669	13000	0.1117	0.7696	0.8218	0.7949	0.9781
0.0082	7.8580	13500	0.1184	0.7821	0.8065	0.7941	0.9786
0.007	8.1490	14000	0.1194	0.7916	0.8039	0.7977	0.9789
0.0068	8.4400	14500	0.1256	0.7858	0.8221	0.8036	0.9789
0.0067	8.7311	15000	0.1244	0.7864	0.8147	0.8003	0.9786
0.0064	9.0221	15500	0.1393	0.7636	0.8227	0.7921	0.9774
0.0049	9.3132	16000	0.1316	0.7727	0.8228	0.7970	0.9783
0.0056	9.6042	16500	0.1245	0.7831	0.8084	0.7955	0.9785
0.005	9.8952	17000	0.1383	0.7605	0.8298	0.7936	0.9781
0.0049	10.1863	17500	0.1368	0.7752	0.8215	0.7977	0.9787
0.0041	10.4773	18000	0.1400	0.7873	0.8155	0.8011	0.9789
0.0049	10.7683	18500	0.1303	0.7896	0.8134	0.8014	0.9788
0.0039	11.0594	19000	0.1434	0.7895	0.8052	0.7973	0.9790
0.0033	11.3504	19500	0.1511	0.7895	0.8162	0.8026	0.9790
0.0039	11.6414	20000	0.1424	0.7749	0.8240	0.7987	0.9787
0.0035	11.9325	20500	0.1528	0.7748	0.8150	0.7944	0.9781
0.0033	12.2235	21000	0.1479	0.7842	0.8191	0.8013	0.9788
0.0028	12.5146	21500	0.1533	0.7745	0.8280	0.8004	0.9786
0.0038	12.8056	22000	0.1417	0.7880	0.8221	0.8047	0.9792
0.0029	13.0966	22500	0.1592	0.7948	0.8121	0.8034	0.9788
0.0025	13.3877	23000	0.1603	0.7721	0.8274	0.7988	0.9783
0.0027	13.6787	23500	0.1516	0.7809	0.8217	0.8008	0.9787
0.0027	13.9697	24000	0.1514	0.7838	0.8237	0.8032	0.9791
0.0023	14.2608	24500	0.1571	0.7842	0.8218	0.8026	0.9789
0.0021	14.5518	25000	0.1666	0.7861	0.8111	0.7984	0.9787
0.0027	14.8428	25500	0.1616	0.7797	0.8285	0.8034	0.9787
0.0023	15.1339	26000	0.1515	0.7876	0.8208	0.8039	0.9789
0.002	15.4249	26500	0.1619	0.7861	0.8090	0.7974	0.9782
0.0021	15.7159	27000	0.1590	0.7857	0.8199	0.8025	0.9789
0.0022	16.0070	27500	0.1593	0.7982	0.8100	0.8041	0.9792
0.0014	16.2980	28000	0.1706	0.7820	0.8176	0.7994	0.9788
0.0017	16.5891	28500	0.1689	0.7854	0.8171	0.8009	0.9791
0.0018	16.8801	29000	0.1623	0.7975	0.8133	0.8053	0.9791
0.0018	17.1711	29500	0.1671	0.7867	0.8217	0.8038	0.9789
0.0014	17.4622	30000	0.1647	0.8033	0.8090	0.8061	0.9794
0.0019	17.7532	30500	0.1651	0.7849	0.8176	0.8009	0.9787
0.0015	18.0442	31000	0.1737	0.7952	0.8146	0.8048	0.9791
0.0011	18.3353	31500	0.1744	0.7820	0.8188	0.8000	0.9786
0.0012	18.6263	32000	0.1696	0.7954	0.8166	0.8059	0.9794
0.0013	18.9173	32500	0.1665	0.7901	0.8253	0.8073	0.9794
0.0011	19.2084	33000	0.1744	0.7833	0.8276	0.8048	0.9789
0.0014	19.4994	33500	0.1748	0.7989	0.8090	0.8039	0.9794
0.0011	19.7905	34000	0.1738	0.7972	0.8101	0.8036	0.9792
0.001	20.0815	34500	0.1768	0.7931	0.8185	0.8056	0.9792
0.0013	20.3725	35000	0.1731	0.7874	0.8211	0.8039	0.9792
0.001	20.6636	35500	0.1756	0.7945	0.8146	0.8044	0.9789
0.001	20.9546	36000	0.1745	0.7957	0.8160	0.8058	0.9792
0.001	21.2456	36500	0.1782	0.7871	0.8165	0.8015	0.9786
0.0007	21.5367	37000	0.1740	0.7929	0.8289	0.8105	0.9799
0.0007	21.8277	37500	0.1811	0.7883	0.8182	0.8030	0.9790
0.0011	22.1187	38000	0.1813	0.8015	0.8106	0.8060	0.9795
0.0008	22.4098	38500	0.1803	0.7840	0.8208	0.8020	0.9786
0.0006	22.7008	39000	0.1786	0.7934	0.8168	0.8049	0.9792
0.0008	22.9919	39500	0.1815	0.7993	0.8171	0.8081	0.9793
0.0007	23.2829	40000	0.1818	0.7956	0.8217	0.8084	0.9793
0.0004	23.5739	40500	0.1901	0.7849	0.8290	0.8063	0.9789
0.001	23.8650	41000	0.1808	0.8008	0.8149	0.8078	0.9795
0.0006	24.1560	41500	0.1821	0.7918	0.8186	0.8050	0.9788
0.0006	24.4470	42000	0.1788	0.8009	0.8153	0.8080	0.9798
0.0007	24.7381	42500	0.1792	0.7942	0.8250	0.8093	0.9795
0.0007	25.0291	43000	0.1781	0.7989	0.8155	0.8071	0.9793
0.0004	25.3201	43500	0.1827	0.7868	0.8250	0.8055	0.9792
0.0005	25.6112	44000	0.1834	0.7953	0.8218	0.8083	0.9795
0.0005	25.9022	44500	0.1849	0.7921	0.8217	0.8066	0.9794
0.0005	26.1932	45000	0.1905	0.7935	0.8228	0.8079	0.9794
0.0003	26.4843	45500	0.1871	0.8001	0.8139	0.8070	0.9794
0.0005	26.7753	46000	0.1864	0.7986	0.8191	0.8087	0.9795
0.0005	27.0664	46500	0.1865	0.7931	0.8215	0.8071	0.9795
0.0003	27.3574	47000	0.1878	0.7925	0.8189	0.8055	0.9794

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1