RoBERTa-ext-lora-chinese-finetuned-ner

This model is a fine-tuned version of hfl/chinese-roberta-wwm-ext on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6506
• Precision: 0.6463
• Recall: 0.7339
• F1: 0.6873
• Accuracy: 0.9081

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: 0.001
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.6585	1.0	126	0.3495	0.4949	0.6045	0.5443	0.8916
0.3145	2.0	252	0.3116	0.5286	0.6644	0.5888	0.9000
0.2644	3.0	378	0.3112	0.5425	0.7012	0.6117	0.9029
0.2373	4.0	504	0.3028	0.5696	0.7090	0.6317	0.9058
0.2078	5.0	630	0.3141	0.5933	0.7102	0.6465	0.9059
0.192	6.0	756	0.3091	0.5842	0.7037	0.6384	0.9069
0.1708	7.0	882	0.3224	0.5803	0.7165	0.6413	0.9046
0.1557	8.0	1008	0.3306	0.6088	0.6833	0.6439	0.9067
0.1424	9.0	1134	0.3243	0.6031	0.6961	0.6463	0.9089
0.1285	10.0	1260	0.3451	0.6041	0.7142	0.6546	0.9064
0.1223	11.0	1386	0.3578	0.6016	0.7080	0.6505	0.9054
0.1111	12.0	1512	0.3615	0.6167	0.7158	0.6625	0.9080
0.1025	13.0	1638	0.3918	0.6073	0.7163	0.6573	0.9051
0.093	14.0	1764	0.3957	0.6119	0.7329	0.6670	0.9078
0.0858	15.0	1890	0.4050	0.6179	0.7052	0.6587	0.9067
0.0769	16.0	2016	0.4218	0.6178	0.7170	0.6637	0.9067
0.0716	17.0	2142	0.4213	0.6090	0.7223	0.6608	0.9057
0.0657	18.0	2268	0.4486	0.6028	0.7299	0.6603	0.9048
0.0652	19.0	2394	0.4388	0.6301	0.7193	0.6718	0.9058
0.0653	20.0	2520	0.4563	0.6165	0.7002	0.6557	0.9039
0.0568	21.0	2646	0.4549	0.6057	0.7283	0.6614	0.9034
0.0522	22.0	2772	0.4711	0.6216	0.7205	0.6674	0.9066
0.0486	23.0	2898	0.4995	0.6267	0.7148	0.6678	0.9062
0.0477	24.0	3024	0.4938	0.6228	0.7261	0.6705	0.9056
0.0415	25.0	3150	0.5129	0.6365	0.7180	0.6748	0.9075
0.0404	26.0	3276	0.5096	0.6287	0.7258	0.6738	0.9072
0.0364	27.0	3402	0.5331	0.6390	0.7205	0.6773	0.9058
0.0362	28.0	3528	0.5572	0.6317	0.7263	0.6757	0.9061
0.0331	29.0	3654	0.5603	0.6377	0.7228	0.6776	0.9050
0.0316	30.0	3780	0.5588	0.6304	0.7256	0.6746	0.9050
0.0321	31.0	3906	0.5579	0.6366	0.7190	0.6753	0.9067
0.0283	32.0	4032	0.5785	0.6469	0.7163	0.6798	0.9067
0.0284	33.0	4158	0.5698	0.6357	0.7246	0.6773	0.9073
0.0256	34.0	4284	0.5816	0.6333	0.7314	0.6788	0.9066
0.0231	35.0	4410	0.6032	0.6273	0.7276	0.6737	0.9059
0.0223	36.0	4536	0.6044	0.6317	0.7238	0.6746	0.9067
0.0225	37.0	4662	0.6007	0.6243	0.7246	0.6707	0.9060
0.0209	38.0	4788	0.6072	0.6325	0.7213	0.6740	0.9067
0.0199	39.0	4914	0.6145	0.6379	0.7261	0.6791	0.9075
0.018	40.0	5040	0.6299	0.6412	0.7341	0.6845	0.9070
0.0174	41.0	5166	0.6264	0.6448	0.7299	0.6847	0.9069
0.0172	42.0	5292	0.6389	0.6409	0.7369	0.6856	0.9061
0.0155	43.0	5418	0.6489	0.6381	0.7241	0.6784	0.9071
0.0156	44.0	5544	0.6418	0.6384	0.7339	0.6828	0.9065
0.0143	45.0	5670	0.6503	0.6378	0.7243	0.6783	0.9066
0.0149	46.0	5796	0.6506	0.6463	0.7339	0.6873	0.9081
0.0135	47.0	5922	0.6497	0.6432	0.7294	0.6836	0.9072
0.0135	48.0	6048	0.6563	0.6389	0.7256	0.6795	0.9069
0.013	49.0	6174	0.6599	0.6377	0.7306	0.6810	0.9066
0.0125	50.0	6300	0.6582	0.6380	0.7278	0.6800	0.9070

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0