scenario-kd-po-ner-full_data-univner_full66

This model is a fine-tuned version of haryoaw/scenario-TCR-NER_data-univner_half on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4139
• Precision: 0.8074
• Recall: 0.7771
• F1: 0.7919
• Accuracy: 0.9789

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
0.9198	0.5828	500	0.6766	0.7412	0.7331	0.7371	0.9745
0.5493	1.1655	1000	0.5975	0.7499	0.7560	0.7529	0.9759
0.4453	1.7483	1500	0.5731	0.7583	0.7585	0.7584	0.9758
0.3762	2.3310	2000	0.5606	0.7824	0.7492	0.7655	0.9764
0.3436	2.9138	2500	0.5208	0.7708	0.7697	0.7703	0.9770
0.3082	3.4965	3000	0.5114	0.7891	0.7491	0.7686	0.9769
0.2933	4.0793	3500	0.5024	0.7873	0.7654	0.7762	0.9774
0.2672	4.6620	4000	0.4971	0.7916	0.7552	0.7729	0.9775
0.2525	5.2448	4500	0.4924	0.7733	0.7775	0.7754	0.9771
0.2385	5.8275	5000	0.4891	0.7833	0.7725	0.7779	0.9775
0.2269	6.4103	5500	0.4843	0.7828	0.7797	0.7813	0.9774
0.2215	6.9930	6000	0.4729	0.7741	0.7862	0.7801	0.9778
0.2076	7.5758	6500	0.4617	0.7838	0.7772	0.7805	0.9780
0.201	8.1585	7000	0.4653	0.7975	0.7671	0.7820	0.9779
0.1935	8.7413	7500	0.4574	0.7785	0.7922	0.7853	0.9778
0.1869	9.3240	8000	0.4662	0.7905	0.7821	0.7863	0.9784
0.1825	9.9068	8500	0.4539	0.7883	0.7807	0.7845	0.9782
0.1748	10.4895	9000	0.4486	0.7975	0.7852	0.7913	0.9789
0.1714	11.0723	9500	0.4499	0.7975	0.7829	0.7901	0.9787
0.166	11.6550	10000	0.4429	0.7931	0.7852	0.7891	0.9787
0.1612	12.2378	10500	0.4427	0.7913	0.7788	0.7850	0.9782
0.1567	12.8205	11000	0.4413	0.8024	0.7762	0.7891	0.9786
0.1544	13.4033	11500	0.4421	0.8068	0.7628	0.7842	0.9781
0.1502	13.9860	12000	0.4388	0.8009	0.7843	0.7925	0.9788
0.146	14.5688	12500	0.4295	0.8	0.7768	0.7882	0.9786
0.1434	15.1515	13000	0.4402	0.8057	0.7755	0.7903	0.9784
0.1404	15.7343	13500	0.4352	0.8106	0.7713	0.7905	0.9785
0.1387	16.3170	14000	0.4360	0.7981	0.7729	0.7853	0.9783
0.1356	16.8998	14500	0.4328	0.8071	0.7722	0.7893	0.9786
0.1345	17.4825	15000	0.4278	0.7990	0.7736	0.7861	0.9786
0.1313	18.0653	15500	0.4268	0.7985	0.7868	0.7926	0.9789
0.1282	18.6480	16000	0.4219	0.7983	0.7818	0.7900	0.9789
0.1284	19.2308	16500	0.4313	0.7968	0.7729	0.7847	0.9782
0.1242	19.8135	17000	0.4255	0.8103	0.7803	0.7950	0.9790
0.1239	20.3963	17500	0.4315	0.8060	0.7720	0.7887	0.9786
0.124	20.9790	18000	0.4317	0.8117	0.7663	0.7883	0.9782
0.1219	21.5618	18500	0.4198	0.7959	0.7758	0.7857	0.9783
0.1199	22.1445	19000	0.4257	0.7976	0.7795	0.7885	0.9784
0.1184	22.7273	19500	0.4271	0.8095	0.7664	0.7874	0.9784
0.118	23.3100	20000	0.4169	0.8076	0.7769	0.7920	0.9789
0.1176	23.8928	20500	0.4203	0.8069	0.7769	0.7916	0.9786
0.1152	24.4755	21000	0.4180	0.8056	0.7816	0.7934	0.9790
0.115	25.0583	21500	0.4206	0.8082	0.7765	0.7921	0.9791
0.1126	25.6410	22000	0.4196	0.8047	0.7762	0.7902	0.9787
0.1148	26.2238	22500	0.4176	0.8061	0.7820	0.7938	0.9789
0.1123	26.8065	23000	0.4156	0.8086	0.7826	0.7954	0.9791
0.1108	27.3893	23500	0.4133	0.8089	0.7829	0.7957	0.9792
0.111	27.9720	24000	0.4114	0.8021	0.7768	0.7893	0.9790
0.1099	28.5548	24500	0.4159	0.8066	0.7739	0.7899	0.9786
0.1112	29.1375	25000	0.4151	0.8082	0.7804	0.7940	0.9789
0.1091	29.7203	25500	0.4139	0.8074	0.7771	0.7919	0.9789

Framework versions

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