scenario-kd-pre-ner-full_data-univner_full66

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4201
• Precision: 0.8240
• Recall: 0.8233
• F1: 0.8236
• Accuracy: 0.9818

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
1.438	0.2910	500	0.8752	0.6871	0.6924	0.6898	0.9698
0.7536	0.5821	1000	0.7361	0.7083	0.7494	0.7283	0.9733
0.6473	0.8731	1500	0.6663	0.7562	0.7471	0.7516	0.9756
0.5663	1.1641	2000	0.6295	0.7463	0.7831	0.7643	0.9768
0.5039	1.4552	2500	0.6083	0.7522	0.7921	0.7716	0.9773
0.4881	1.7462	3000	0.5952	0.7849	0.7712	0.7780	0.9780
0.4565	2.0373	3500	0.5658	0.7951	0.7876	0.7913	0.9786
0.4024	2.3283	4000	0.5517	0.7801	0.7850	0.7825	0.9783
0.3991	2.6193	4500	0.5389	0.7833	0.8057	0.7943	0.9795
0.3865	2.9104	5000	0.5337	0.7950	0.7953	0.7952	0.9797
0.3492	3.2014	5500	0.5218	0.7965	0.8015	0.7990	0.9796
0.3414	3.4924	6000	0.5356	0.7881	0.7960	0.7920	0.9792
0.3288	3.7835	6500	0.5314	0.7830	0.8048	0.7937	0.9792
0.3269	4.0745	7000	0.5114	0.7821	0.8046	0.7932	0.9796
0.2907	4.3655	7500	0.5038	0.7939	0.8035	0.7987	0.9799
0.2974	4.6566	8000	0.5001	0.7938	0.8147	0.8041	0.9801
0.2941	4.9476	8500	0.5036	0.8036	0.7960	0.7998	0.9797
0.269	5.2386	9000	0.4910	0.8089	0.8000	0.8044	0.9802
0.2602	5.5297	9500	0.4833	0.7993	0.8117	0.8054	0.9800
0.2681	5.8207	10000	0.4800	0.8029	0.8121	0.8075	0.9799
0.2572	6.1118	10500	0.4757	0.7987	0.8090	0.8038	0.9801
0.2415	6.4028	11000	0.4825	0.7998	0.8116	0.8056	0.9801
0.2393	6.6938	11500	0.4786	0.7963	0.8149	0.8055	0.9801
0.241	6.9849	12000	0.4760	0.8068	0.7992	0.8030	0.9798
0.2215	7.2759	12500	0.4685	0.8046	0.8103	0.8074	0.9805
0.2233	7.5669	13000	0.4703	0.8117	0.8029	0.8073	0.9802
0.2197	7.8580	13500	0.4620	0.8024	0.8208	0.8115	0.9803
0.2153	8.1490	14000	0.4776	0.8192	0.7904	0.8045	0.9801
0.2045	8.4400	14500	0.4675	0.8174	0.8127	0.8151	0.9809
0.2044	8.7311	15000	0.4664	0.8180	0.8032	0.8105	0.9808
0.2058	9.0221	15500	0.4543	0.8070	0.8107	0.8088	0.9805
0.1937	9.3132	16000	0.4604	0.8040	0.8263	0.8150	0.9809
0.1921	9.6042	16500	0.4590	0.8138	0.8228	0.8183	0.9808
0.19	9.8952	17000	0.4550	0.8101	0.8145	0.8123	0.9809
0.1863	10.1863	17500	0.4568	0.8112	0.8074	0.8093	0.9806
0.1806	10.4773	18000	0.4487	0.8183	0.8123	0.8153	0.9808
0.1791	10.7683	18500	0.4534	0.8147	0.8113	0.8130	0.9809
0.1768	11.0594	19000	0.4502	0.8040	0.8192	0.8115	0.9809
0.1712	11.3504	19500	0.4553	0.8126	0.8184	0.8155	0.9805
0.1699	11.6414	20000	0.4520	0.8101	0.8173	0.8137	0.9811
0.1731	11.9325	20500	0.4530	0.8065	0.8098	0.8081	0.9807
0.1673	12.2235	21000	0.4524	0.8209	0.8064	0.8136	0.9808
0.1627	12.5146	21500	0.4495	0.8210	0.8133	0.8171	0.9810
0.1653	12.8056	22000	0.4590	0.8210	0.7986	0.8096	0.9804
0.1609	13.0966	22500	0.4542	0.8202	0.8153	0.8177	0.9808
0.1555	13.3877	23000	0.4528	0.8194	0.8121	0.8157	0.9809
0.1566	13.6787	23500	0.4473	0.8169	0.8149	0.8159	0.9808
0.1572	13.9697	24000	0.4433	0.8259	0.8071	0.8164	0.9810
0.1502	14.2608	24500	0.4484	0.8285	0.8093	0.8188	0.9813
0.1494	14.5518	25000	0.4509	0.8139	0.8146	0.8142	0.9807
0.1504	14.8428	25500	0.4490	0.8199	0.8166	0.8183	0.9812
0.1477	15.1339	26000	0.4318	0.8255	0.8160	0.8207	0.9815
0.1431	15.4249	26500	0.4372	0.8183	0.8176	0.8180	0.9813
0.145	15.7159	27000	0.4367	0.8147	0.8139	0.8143	0.9812
0.1437	16.0070	27500	0.4340	0.8207	0.8222	0.8215	0.9816
0.1387	16.2980	28000	0.4454	0.8254	0.8098	0.8176	0.9811
0.1396	16.5891	28500	0.4424	0.8238	0.8120	0.8178	0.9813
0.1399	16.8801	29000	0.4343	0.8171	0.8172	0.8171	0.9812
0.1362	17.1711	29500	0.4425	0.8222	0.8162	0.8192	0.9812
0.1365	17.4622	30000	0.4295	0.8238	0.8186	0.8212	0.9813
0.134	17.7532	30500	0.4312	0.8234	0.8153	0.8193	0.9815
0.1342	18.0442	31000	0.4268	0.8240	0.8214	0.8227	0.9816
0.1308	18.3353	31500	0.4393	0.8314	0.8098	0.8205	0.9812
0.131	18.6263	32000	0.4323	0.8300	0.8194	0.8247	0.9817
0.1304	18.9173	32500	0.4390	0.8216	0.8140	0.8178	0.9812
0.1287	19.2084	33000	0.4297	0.8275	0.8114	0.8194	0.9814
0.1287	19.4994	33500	0.4323	0.8260	0.8123	0.8191	0.9813
0.1279	19.7905	34000	0.4216	0.8185	0.8222	0.8204	0.9814
0.1273	20.0815	34500	0.4224	0.8270	0.8163	0.8216	0.9817
0.1253	20.3725	35000	0.4251	0.8294	0.8181	0.8237	0.9814
0.1249	20.6636	35500	0.4214	0.8274	0.8204	0.8239	0.9818
0.124	20.9546	36000	0.4331	0.8282	0.8149	0.8215	0.9813
0.122	21.2456	36500	0.4250	0.8274	0.8159	0.8216	0.9814
0.1224	21.5367	37000	0.4297	0.8316	0.8150	0.8232	0.9813
0.1226	21.8277	37500	0.4242	0.8251	0.8231	0.8241	0.9817
0.1206	22.1187	38000	0.4269	0.8285	0.8142	0.8213	0.9813
0.1188	22.4098	38500	0.4226	0.8247	0.8201	0.8224	0.9814
0.1202	22.7008	39000	0.4177	0.8235	0.8195	0.8215	0.9815
0.1199	22.9919	39500	0.4272	0.8268	0.8111	0.8189	0.9814
0.1184	23.2829	40000	0.4264	0.8298	0.8116	0.8206	0.9814
0.1168	23.5739	40500	0.4283	0.8267	0.8160	0.8213	0.9815
0.1177	23.8650	41000	0.4221	0.8312	0.8094	0.8202	0.9814
0.1177	24.1560	41500	0.4150	0.8233	0.8220	0.8226	0.9818
0.1155	24.4470	42000	0.4201	0.8240	0.8233	0.8236	0.9818

Framework versions

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