README.md

metadata

base_model: bert-base-chinese
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: classifier_adapter
    results: []

classifier_adapter

This model is a fine-tuned version of bert-base-chinese on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0386
• Accuracy: 0.9875
• Precision: 0.8841
• Recall: 0.7947
• F1: 0.8283
• Ap: 0.8850

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 0
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Ap
No log	0.38	100	0.1590	0.9571	0.0	0.0	0.0	0.1046
No log	0.75	200	0.1578	0.9571	0.0	0.0	0.0	0.1808
No log	1.13	300	0.1185	0.9653	0.0899	0.0599	0.0680	0.4391
No log	1.51	400	0.0898	0.9724	0.2199	0.1409	0.1617	0.6479
0.1405	1.89	500	0.0774	0.9750	0.3319	0.2273	0.2575	0.7417
0.1405	2.26	600	0.0683	0.9771	0.4118	0.3002	0.3294	0.7791
0.1405	2.64	700	0.0616	0.9804	0.6207	0.4336	0.4810	0.8187
0.1405	3.02	800	0.0556	0.9821	0.7210	0.4875	0.5435	0.8380
0.1405	3.4	900	0.0519	0.9830	0.7329	0.5224	0.5839	0.8566
0.0598	3.77	1000	0.0486	0.9846	0.7818	0.6063	0.6615	0.8629
0.0598	4.15	1100	0.0469	0.9853	0.8223	0.6807	0.7248	0.8633
0.0598	4.53	1200	0.0457	0.9856	0.8521	0.7235	0.7663	0.8666
0.0598	4.91	1300	0.0439	0.9859	0.8436	0.6955	0.7435	0.8753
0.0598	5.28	1400	0.0424	0.9862	0.8715	0.6964	0.7496	0.8739
0.0399	5.66	1500	0.0415	0.9869	0.8695	0.7621	0.7994	0.8772
0.0399	6.04	1600	0.0416	0.9865	0.8700	0.7670	0.8039	0.8853
0.0399	6.42	1700	0.0401	0.9871	0.8687	0.7686	0.8047	0.8846
0.0399	6.79	1800	0.0405	0.9867	0.8734	0.7851	0.8167	0.8848
0.0399	7.17	1900	0.0410	0.9865	0.8600	0.7708	0.8057	0.8770
0.0315	7.55	2000	0.0393	0.9873	0.8869	0.7718	0.8158	0.8819
0.0315	7.92	2100	0.0385	0.9871	0.8747	0.7861	0.8196	0.8856
0.0315	8.3	2200	0.0386	0.9877	0.8863	0.7856	0.8227	0.8857
0.0315	8.68	2300	0.0390	0.9869	0.8695	0.7949	0.8221	0.8830
0.0315	9.06	2400	0.0391	0.9872	0.8685	0.8081	0.8311	0.8830
0.026	9.43	2500	0.0386	0.9875	0.8841	0.7947	0.8283	0.8850
0.026	9.81	2600	0.0390	0.9871	0.8615	0.8064	0.8264	0.8840
0.026	10.19	2700	0.0386	0.9873	0.8689	0.8023	0.8264	0.8859
0.026	10.57	2800	0.0386	0.9873	0.8737	0.7986	0.8265	0.8860

Framework versions

• Transformers 4.36.2
• Pytorch 2.2.1+cu121
• Tokenizers 0.15.2