scenario-non-kd-pre-ner-half-mdeberta_data-univner_full66

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: 0.1673
• Precision: 0.6897
• Recall: 0.7416
• F1: 0.7147
• Accuracy: 0.9712

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: 66
• 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
0.3169	0.29	500	0.2023	0.2870	0.3272	0.3058	0.9356
0.1897	0.58	1000	0.1647	0.4142	0.4800	0.4447	0.9459
0.1557	0.87	1500	0.1464	0.4614	0.5513	0.5023	0.9516
0.1336	1.16	2000	0.1368	0.4959	0.5706	0.5307	0.9543
0.117	1.46	2500	0.1359	0.5037	0.6188	0.5554	0.9541
0.1133	1.75	3000	0.1230	0.5571	0.6091	0.5820	0.9601
0.1062	2.04	3500	0.1214	0.5613	0.6682	0.6101	0.9609
0.0905	2.33	4000	0.1213	0.5726	0.6721	0.6183	0.9614
0.0884	2.62	4500	0.1115	0.6075	0.6738	0.6389	0.9649
0.0862	2.91	5000	0.1202	0.5708	0.7003	0.6290	0.9604
0.0765	3.2	5500	0.1148	0.5965	0.7125	0.6494	0.9644
0.0728	3.49	6000	0.1109	0.6109	0.7052	0.6547	0.9654
0.0706	3.78	6500	0.1164	0.6072	0.7031	0.6516	0.9641
0.0699	4.07	7000	0.1121	0.6404	0.7058	0.6715	0.9671
0.0615	4.37	7500	0.1140	0.6343	0.7143	0.6719	0.9670
0.0595	4.66	8000	0.1128	0.6284	0.7192	0.6707	0.9665
0.0603	4.95	8500	0.1113	0.6368	0.7267	0.6788	0.9676
0.0534	5.24	9000	0.1134	0.6662	0.7112	0.6879	0.9693
0.0522	5.53	9500	0.1168	0.6282	0.7298	0.6752	0.9665
0.049	5.82	10000	0.1157	0.6429	0.7250	0.6815	0.9682
0.0488	6.11	10500	0.1170	0.6660	0.7188	0.6914	0.9689
0.0434	6.4	11000	0.1176	0.6701	0.7013	0.6854	0.9682
0.0449	6.69	11500	0.1183	0.6474	0.7236	0.6834	0.9672
0.0437	6.98	12000	0.1208	0.6374	0.7324	0.6816	0.9674
0.0395	7.28	12500	0.1192	0.6722	0.7042	0.6879	0.9690
0.0385	7.57	13000	0.1207	0.6558	0.7251	0.6887	0.9695
0.0389	7.86	13500	0.1168	0.6850	0.7146	0.6995	0.9699
0.0336	8.15	14000	0.1281	0.6509	0.7329	0.6895	0.9685
0.0324	8.44	14500	0.1297	0.6414	0.7394	0.6870	0.9669
0.0318	8.73	15000	0.1269	0.6658	0.7321	0.6974	0.9698
0.034	9.02	15500	0.1223	0.6783	0.7165	0.6969	0.9701
0.0291	9.31	16000	0.1301	0.6797	0.7217	0.7001	0.9698
0.0296	9.6	16500	0.1293	0.6849	0.7127	0.6985	0.9698
0.0288	9.9	17000	0.1286	0.6566	0.7322	0.6924	0.9693
0.0274	10.19	17500	0.1323	0.6721	0.7201	0.6953	0.9689
0.0259	10.48	18000	0.1310	0.6782	0.7296	0.7029	0.9699
0.0252	10.77	18500	0.1350	0.6746	0.7202	0.6967	0.9701
0.0246	11.06	19000	0.1347	0.6740	0.7384	0.7047	0.9701
0.0214	11.35	19500	0.1399	0.6736	0.7347	0.7028	0.9704
0.0227	11.64	20000	0.1377	0.6730	0.7358	0.7030	0.9695
0.0227	11.93	20500	0.1365	0.6994	0.7205	0.7098	0.9714
0.0213	12.22	21000	0.1351	0.6772	0.7363	0.7055	0.9703
0.0191	12.51	21500	0.1407	0.6702	0.7465	0.7063	0.9699
0.0198	12.81	22000	0.1403	0.6829	0.7237	0.7027	0.9701
0.0197	13.1	22500	0.1418	0.6820	0.7351	0.7075	0.9707
0.0184	13.39	23000	0.1387	0.7009	0.7267	0.7136	0.9712
0.0173	13.68	23500	0.1471	0.6629	0.7381	0.6985	0.9692
0.018	13.97	24000	0.1470	0.6786	0.7306	0.7037	0.9698
0.016	14.26	24500	0.1493	0.6832	0.7350	0.7081	0.9704
0.0159	14.55	25000	0.1508	0.6715	0.7533	0.7101	0.9697
0.0158	14.84	25500	0.1526	0.6792	0.7361	0.7065	0.9694
0.0155	15.13	26000	0.1491	0.6936	0.7393	0.7157	0.9713
0.014	15.42	26500	0.1523	0.6975	0.7376	0.7170	0.9713
0.0145	15.72	27000	0.1503	0.7004	0.7227	0.7114	0.9709
0.0142	16.01	27500	0.1533	0.6804	0.7442	0.7109	0.9705
0.0131	16.3	28000	0.1533	0.6866	0.7403	0.7124	0.9706
0.0129	16.59	28500	0.1549	0.6798	0.7363	0.7069	0.9697
0.0128	16.88	29000	0.1591	0.6902	0.7314	0.7102	0.9710
0.0121	17.17	29500	0.1596	0.6780	0.7399	0.7076	0.9703
0.0116	17.46	30000	0.1576	0.6929	0.7347	0.7132	0.9710
0.012	17.75	30500	0.1546	0.6888	0.7370	0.7121	0.9708
0.0126	18.04	31000	0.1619	0.6927	0.7361	0.7138	0.9710
0.0112	18.34	31500	0.1610	0.6809	0.7415	0.7099	0.9702
0.011	18.63	32000	0.1603	0.6873	0.7319	0.7089	0.9706
0.011	18.92	32500	0.1644	0.6743	0.7389	0.7051	0.9696
0.0102	19.21	33000	0.1625	0.6966	0.7342	0.7149	0.9711
0.0098	19.5	33500	0.1678	0.6993	0.7311	0.7148	0.9711
0.0103	19.79	34000	0.1657	0.6911	0.7410	0.7152	0.9709
0.0102	20.08	34500	0.1649	0.6954	0.7275	0.7110	0.9712
0.0095	20.37	35000	0.1683	0.6957	0.7266	0.7108	0.9708
0.0095	20.66	35500	0.1661	0.6914	0.7344	0.7122	0.9710
0.0089	20.95	36000	0.1678	0.6962	0.7302	0.7128	0.9713
0.0084	21.25	36500	0.1673	0.6897	0.7416	0.7147	0.9712

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3