scenario-kd-pre-ner-full_data-univner_full44

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: 0.4193
• Precision: 0.8281
• Recall: 0.8175
• F1: 0.8228
• Accuracy: 0.9817

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
1.4887	0.2910	500	0.8780	0.6630	0.6943	0.6783	0.9693
0.7434	0.5821	1000	0.7247	0.7040	0.7534	0.7279	0.9735
0.6458	0.8731	1500	0.6695	0.7248	0.7692	0.7463	0.9752
0.562	1.1641	2000	0.6301	0.7474	0.7830	0.7648	0.9768
0.5142	1.4552	2500	0.6432	0.7960	0.7413	0.7677	0.9773
0.4901	1.7462	3000	0.5939	0.7664	0.7689	0.7676	0.9768
0.4625	2.0373	3500	0.5650	0.7763	0.7899	0.7830	0.9784
0.4058	2.3283	4000	0.5634	0.7941	0.7807	0.7873	0.9786
0.3987	2.6193	4500	0.5539	0.7763	0.8000	0.7880	0.9789
0.3879	2.9104	5000	0.5344	0.7848	0.8042	0.7944	0.9789
0.3586	3.2014	5500	0.5491	0.7907	0.7915	0.7911	0.9791
0.3418	3.4924	6000	0.5209	0.7757	0.8121	0.7935	0.9789
0.3334	3.7835	6500	0.5221	0.7954	0.8000	0.7977	0.9797
0.3254	4.0745	7000	0.5208	0.8027	0.7943	0.7985	0.9793
0.3037	4.3655	7500	0.5119	0.7912	0.8000	0.7956	0.9797
0.2948	4.6566	8000	0.5057	0.7966	0.8062	0.8014	0.9796
0.2929	4.9476	8500	0.5024	0.8051	0.7987	0.8019	0.9796
0.2721	5.2386	9000	0.5030	0.8024	0.7868	0.7945	0.9796
0.2654	5.5297	9500	0.4919	0.8124	0.7940	0.8031	0.9800
0.2664	5.8207	10000	0.4992	0.7986	0.8121	0.8053	0.9798
0.2582	6.1118	10500	0.4874	0.8126	0.8000	0.8063	0.9802
0.2404	6.4028	11000	0.4980	0.8081	0.8000	0.8040	0.9803
0.2408	6.6938	11500	0.4875	0.8026	0.8036	0.8031	0.9800
0.2416	6.9849	12000	0.4830	0.8074	0.7982	0.8027	0.9799
0.2246	7.2759	12500	0.4750	0.8084	0.8116	0.8100	0.9805
0.2225	7.5669	13000	0.4839	0.8017	0.8162	0.8089	0.9807
0.2235	7.8580	13500	0.4676	0.8052	0.8134	0.8093	0.9807
0.2111	8.1490	14000	0.4718	0.8151	0.8022	0.8086	0.9806
0.207	8.4400	14500	0.4777	0.8036	0.8165	0.8100	0.9802
0.2063	8.7311	15000	0.4704	0.8250	0.7995	0.8120	0.9806
0.2051	9.0221	15500	0.4718	0.8027	0.8119	0.8073	0.9803
0.1922	9.3132	16000	0.4767	0.8154	0.8077	0.8115	0.9806
0.192	9.6042	16500	0.4735	0.8160	0.8114	0.8137	0.9811
0.1946	9.8952	17000	0.4711	0.8100	0.8176	0.8138	0.9807
0.1843	10.1863	17500	0.4666	0.8096	0.8113	0.8105	0.9808
0.1801	10.4773	18000	0.4606	0.8064	0.8150	0.8107	0.9805
0.1824	10.7683	18500	0.4573	0.8158	0.8133	0.8145	0.9810
0.1775	11.0594	19000	0.4733	0.8209	0.7951	0.8078	0.9803
0.1723	11.3504	19500	0.4567	0.8164	0.8168	0.8166	0.9813
0.1716	11.6414	20000	0.4596	0.8153	0.8062	0.8107	0.9809
0.1696	11.9325	20500	0.4553	0.8141	0.8250	0.8195	0.9813
0.165	12.2235	21000	0.4474	0.8225	0.8114	0.8169	0.9810
0.1609	12.5146	21500	0.4638	0.8189	0.8094	0.8141	0.9810
0.1648	12.8056	22000	0.4459	0.8122	0.8120	0.8121	0.9809
0.1599	13.0966	22500	0.4509	0.8184	0.8104	0.8144	0.9811
0.1556	13.3877	23000	0.4603	0.8167	0.8062	0.8114	0.9808
0.1559	13.6787	23500	0.4515	0.8163	0.8150	0.8157	0.9807
0.1546	13.9697	24000	0.4436	0.8089	0.8225	0.8157	0.9809
0.1487	14.2608	24500	0.4422	0.8114	0.8228	0.8170	0.9811
0.1503	14.5518	25000	0.4467	0.8180	0.8169	0.8174	0.9813
0.1485	14.8428	25500	0.4508	0.8098	0.8215	0.8156	0.9807
0.1466	15.1339	26000	0.4441	0.8157	0.8147	0.8152	0.9812
0.1432	15.4249	26500	0.4473	0.8242	0.8111	0.8176	0.9813
0.1431	15.7159	27000	0.4513	0.8194	0.8159	0.8177	0.9812
0.1444	16.0070	27500	0.4381	0.8166	0.8209	0.8188	0.9812
0.1373	16.2980	28000	0.4420	0.8163	0.8234	0.8199	0.9815
0.1375	16.5891	28500	0.4395	0.8203	0.8179	0.8191	0.9815
0.1405	16.8801	29000	0.4409	0.8227	0.8126	0.8176	0.9810
0.1369	17.1711	29500	0.4371	0.8259	0.8124	0.8191	0.9811
0.1345	17.4622	30000	0.4428	0.8248	0.8096	0.8171	0.9809
0.1356	17.7532	30500	0.4341	0.8275	0.8175	0.8225	0.9815
0.1323	18.0442	31000	0.4312	0.8229	0.8199	0.8214	0.9813
0.1302	18.3353	31500	0.4308	0.8242	0.8222	0.8232	0.9819
0.1302	18.6263	32000	0.4308	0.8217	0.8159	0.8188	0.9814
0.1308	18.9173	32500	0.4371	0.8274	0.8042	0.8156	0.9813
0.1289	19.2084	33000	0.4339	0.8305	0.8108	0.8206	0.9815
0.1273	19.4994	33500	0.4358	0.8176	0.8158	0.8167	0.9813
0.1271	19.7905	34000	0.4403	0.8229	0.8123	0.8175	0.9810
0.125	20.0815	34500	0.4280	0.8235	0.8201	0.8218	0.9815
0.1259	20.3725	35000	0.4341	0.8243	0.8124	0.8183	0.9812
0.1233	20.6636	35500	0.4327	0.8282	0.8075	0.8177	0.9812
0.1243	20.9546	36000	0.4253	0.8252	0.8192	0.8222	0.9814
0.1233	21.2456	36500	0.4333	0.8203	0.8114	0.8158	0.9812
0.1202	21.5367	37000	0.4253	0.8196	0.8168	0.8182	0.9814
0.1223	21.8277	37500	0.4234	0.8311	0.8142	0.8225	0.9815
0.1215	22.1187	38000	0.4203	0.8249	0.8197	0.8223	0.9818
0.1177	22.4098	38500	0.4200	0.8280	0.8225	0.8253	0.9818
0.1198	22.7008	39000	0.4257	0.8267	0.8199	0.8233	0.9818
0.1187	22.9919	39500	0.4253	0.8274	0.8222	0.8248	0.9817
0.1179	23.2829	40000	0.4261	0.8267	0.8163	0.8215	0.9812
0.1168	23.5739	40500	0.4203	0.8295	0.8156	0.8225	0.9815
0.1174	23.8650	41000	0.4216	0.8278	0.8145	0.8211	0.9816
0.1159	24.1560	41500	0.4226	0.8271	0.8207	0.8239	0.9818
0.1147	24.4470	42000	0.4274	0.8328	0.8147	0.8237	0.9814
0.1168	24.7381	42500	0.4240	0.8221	0.8147	0.8184	0.9815
0.1148	25.0291	43000	0.4222	0.8224	0.8119	0.8171	0.9814
0.1141	25.3201	43500	0.4179	0.8248	0.8197	0.8222	0.9818
0.1142	25.6112	44000	0.4204	0.8235	0.8178	0.8206	0.9815
0.1131	25.9022	44500	0.4190	0.8342	0.8222	0.8282	0.9818
0.114	26.1932	45000	0.4247	0.8289	0.8201	0.8245	0.9816
0.1119	26.4843	45500	0.4198	0.8290	0.8179	0.8234	0.9815
0.1132	26.7753	46000	0.4221	0.8224	0.8166	0.8195	0.9814
0.1125	27.0664	46500	0.4216	0.8306	0.8129	0.8216	0.9814
0.1103	27.3574	47000	0.4232	0.8260	0.8126	0.8193	0.9813
0.1113	27.6484	47500	0.4200	0.8321	0.8150	0.8235	0.9815
0.112	27.9395	48000	0.4186	0.8285	0.8227	0.8256	0.9817
0.111	28.2305	48500	0.4203	0.8326	0.8182	0.8254	0.9817
0.1095	28.5215	49000	0.4194	0.8300	0.8173	0.8236	0.9816
0.1104	28.8126	49500	0.4212	0.8246	0.8192	0.8219	0.9815
0.1098	29.1036	50000	0.4189	0.8278	0.8165	0.8221	0.9814
0.1097	29.3946	50500	0.4176	0.8322	0.8173	0.8247	0.9817
0.1098	29.6857	51000	0.4173	0.8252	0.8165	0.8208	0.9813
0.1104	29.9767	51500	0.4193	0.8281	0.8175	0.8228	0.9817

Framework versions

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