jetaudio/chinese-novels-ner

chinese-electra-180g-base-discriminator-pro-ner-final

This model is a fine-tuned version of hfl/chinese-electra-180g-base-discriminator on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0940
• Overall Precision: 0.8057
• Overall Recall: 0.8460
• Overall F1: 0.8253
• Overall Accuracy: 0.9698
• Ucm: 0.7474
• Lcm: 0.7140

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: reduce_lr_on_plateau
• lr_scheduler_warmup_steps: 100
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Overall Precision	Overall Recall	Overall F1	Overall Accuracy	Ucm	Lcm
0.3089	0.0420	1000	0.2055	0.6352	0.6655	0.6500	0.9457	0.6555	0.6075
0.187	0.0840	2000	0.1547	0.6632	0.7649	0.7104	0.9537	0.6670	0.6200
0.1208	0.1260	3000	0.1396	0.7391	0.7471	0.7431	0.9575	0.6775	0.6409
0.0619	0.1680	4000	0.1323	0.7548	0.7713	0.7629	0.9614	0.7098	0.6639
0.153	0.2100	5000	0.1231	0.7545	0.8144	0.7833	0.9629	0.7244	0.6785
0.2363	0.2520	6000	0.1252	0.7424	0.8299	0.7837	0.9624	0.7182	0.6722
0.0843	0.2940	7000	0.1191	0.7625	0.8046	0.7830	0.9629	0.7046	0.6733
0.2406	0.3360	8000	0.1156	0.7785	0.8040	0.7911	0.9651	0.7213	0.6879
0.1422	0.3780	9000	0.1156	0.7791	0.8109	0.7947	0.9647	0.7150	0.6806
0.1183	0.4200	10000	0.1107	0.7634	0.8287	0.7947	0.9652	0.7223	0.6816
0.1386	0.4620	11000	0.1095	0.7796	0.8293	0.8037	0.9654	0.7223	0.6806
0.1275	0.5040	12000	0.1062	0.7865	0.8425	0.8135	0.9671	0.7432	0.7004
0.1585	0.5460	13000	0.0997	0.7992	0.8420	0.8200	0.9688	0.7495	0.7088
0.0954	0.5880	14000	0.0973	0.7997	0.8282	0.8137	0.9686	0.7422	0.7077
0.1024	0.6300	15000	0.0964	0.8004	0.8414	0.8204	0.9698	0.7380	0.7035
0.116	0.6720	16000	0.0954	0.7917	0.8368	0.8136	0.9698	0.7317	0.7025
0.1147	0.7140	17000	0.0959	0.8043	0.8552	0.8290	0.9696	0.7474	0.7150
0.0163	0.7560	18000	0.0953	0.8032	0.8397	0.8210	0.9700	0.7474	0.7161
0.0926	0.7980	19000	0.0949	0.8079	0.8362	0.8218	0.9703	0.7463	0.7150
0.1387	0.8400	20000	0.0944	0.7972	0.8448	0.8203	0.9698	0.7463	0.7150
0.1732	0.8820	21000	0.0932	0.8120	0.8489	0.8300	0.9703	0.7537	0.7213
0.1174	0.9240	22000	0.0930	0.8141	0.8408	0.8273	0.9703	0.7463	0.7171
0.0835	0.9660	23000	0.0935	0.8033	0.8448	0.8235	0.9694	0.7516	0.7192
0.0353	1.0080	24000	0.0930	0.8109	0.8454	0.8278	0.9699	0.7547	0.7213
0.0628	1.0500	25000	0.0931	0.8076	0.8420	0.8244	0.9701	0.7505	0.7182
0.0683	1.0920	26000	0.0935	0.8075	0.8460	0.8263	0.9699	0.7516	0.7182
0.0178	1.1340	27000	0.0938	0.8054	0.8466	0.8254	0.9700	0.7495	0.7161
0.0758	1.1760	28000	0.0937	0.8106	0.8460	0.8279	0.9700	0.7526	0.7182
0.1645	1.2180	29000	0.0937	0.8054	0.8443	0.8244	0.9701	0.7484	0.7161
0.109	1.2600	30000	0.0941	0.8107	0.8466	0.8282	0.9701	0.7505	0.7182
0.057	1.3020	31000	0.0941	0.8058	0.8466	0.8257	0.9696	0.7474	0.7140
0.0791	1.3440	32000	0.0940	0.8101	0.8483	0.8287	0.9699	0.7495	0.7182
0.0186	1.3860	33000	0.0940	0.8087	0.8477	0.8277	0.9698	0.7495	0.7161
0.0606	1.4280	34000	0.0940	0.8103	0.8471	0.8283	0.9700	0.7505	0.7171
0.0183	1.4700	35000	0.0940	0.8089	0.8466	0.8273	0.9700	0.7495	0.7161
0.0213	1.5120	36000	0.0940	0.8108	0.8471	0.8286	0.9700	0.7516	0.7171
0.0241	1.5540	37000	0.0941	0.8098	0.8466	0.8278	0.9698	0.7505	0.7161
0.0502	1.5960	38000	0.0941	0.8089	0.8466	0.8273	0.9699	0.7495	0.7161
0.2536	1.6380	39000	0.0941	0.8080	0.8466	0.8268	0.9698	0.7484	0.7150
0.021	1.6800	40000	0.0941	0.8076	0.8466	0.8266	0.9698	0.7484	0.7150
0.2905	1.7220	41000	0.0941	0.8080	0.8466	0.8268	0.9698	0.7484	0.7150
0.0292	1.7640	42000	0.0941	0.8080	0.8466	0.8268	0.9699	0.7495	0.7161
0.0864	1.8060	43000	0.0940	0.8080	0.8466	0.8268	0.9699	0.7484	0.7150
0.0233	1.8480	44000	0.0940	0.8070	0.8460	0.8260	0.9698	0.7474	0.7140
0.0813	1.8900	45000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0481	1.9320	46000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.124	1.9740	47000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0534	2.0160	48000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0647	2.0580	49000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.104	2.1000	50000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0765	2.1420	51000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0882	2.1840	52000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0778	2.2260	53000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1016	2.2680	54000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0329	2.3101	55000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1606	2.3521	56000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1131	2.3941	57000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.078	2.4361	58000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0388	2.4781	59000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1212	2.5201	60000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0687	2.5621	61000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0633	2.6041	62000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0986	2.6461	63000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1169	2.6881	64000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0476	2.7301	65000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0126	2.7721	66000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0185	2.8141	67000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0491	2.8561	68000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.0996	2.8981	69000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1328	2.9401	70000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140
0.1135	2.9821	71000	0.0940	0.8057	0.8460	0.8253	0.9698	0.7474	0.7140

Framework versions

• Transformers 4.53.1
• Pytorch 2.6.0+cu126
• Datasets 3.5.1
• Tokenizers 0.21.1