Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Hamming
0.31	0.02	1000	0.0780	0.9902	0.0	0.0	0.0	0.0098
0.0581	0.03	2000	0.0503	0.9902	0.0	0.0	0.0	0.0098
0.0488	0.05	3000	0.0459	0.9902	0.0	0.0	0.0	0.0098
0.0407	0.06	4000	0.0355	0.9911	0.8475	0.1130	0.1994	0.0089
0.0323	0.08	5000	0.0293	0.9918	0.8327	0.1961	0.3175	0.0082
0.0278	0.09	6000	0.0255	0.9924	0.8223	0.2846	0.4228	0.0076
0.0246	0.11	7000	0.0235	0.9929	0.8002	0.3609	0.4974	0.0071
0.0231	0.13	8000	0.0218	0.9933	0.7988	0.4189	0.5496	0.0067
0.0217	0.14	9000	0.0209	0.9934	0.7888	0.4444	0.5685	0.0066
0.0208	0.16	10000	0.0201	0.9935	0.8036	0.4473	0.5747	0.0065
0.0199	0.17	11000	0.0195	0.9936	0.7901	0.4751	0.5934	0.0064
0.0195	0.19	12000	0.0192	0.9938	0.7889	0.4923	0.6063	0.0062
0.019	0.21	13000	0.0188	0.9938	0.7999	0.4913	0.6087	0.0062
0.0189	0.22	14000	0.0184	0.9939	0.7872	0.5094	0.6186	0.0061
0.0188	0.24	15000	0.0182	0.9939	0.7920	0.5084	0.6193	0.0061
0.0183	0.25	16000	0.0180	0.9940	0.7901	0.5241	0.6301	0.0060
0.0181	0.27	17000	0.0179	0.9940	0.7897	0.5277	0.6327	0.0060
0.0179	0.28	18000	0.0177	0.9941	0.7928	0.5301	0.6353	0.0059
0.0179	0.3	19000	0.0176	0.9941	0.7953	0.5275	0.6343	0.0059
0.0178	0.32	20000	0.0176	0.9941	0.7976	0.5324	0.6386	0.0059

bdpc/SciBERT_20K_steps

SciBERT_20K_steps