bert-base-chinese-finetuned-split

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: 0.0154
• F1: 1.0

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: 5e-05
• train_batch_size: 27
• eval_batch_size: 27
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1
3.4938	1.0	1	1.6750	0.0
1.7935	2.0	2	0.7116	0.0
0.7945	3.0	3	0.5122	0.0
0.5331	4.0	4	0.4022	0.0
0.4232	5.0	5	0.3155	0.0
0.3449	6.0	6	0.2706	0.0
0.2951	7.0	7	0.2177	0.1481
0.2455	8.0	8	0.1809	0.2143
0.2115	9.0	9	0.1463	0.6486
0.1631	10.0	10	0.1210	0.8636
0.1427	11.0	11	0.0988	0.8889
0.1227	12.0	12	0.0736	0.9583
0.098	13.0	13	0.0528	0.9796
0.0769	14.0	14	0.0399	1.0
0.0616	15.0	15	0.0319	1.0
0.0526	16.0	16	0.0267	1.0
0.0505	17.0	17	0.0216	1.0
0.0337	18.0	18	0.0182	1.0
0.034	19.0	19	0.0163	1.0
0.0275	20.0	20	0.0154	1.0

Framework versions

• Transformers 4.18.0
• Pytorch 1.12.0+cu102
• Datasets 1.18.4
• Tokenizers 0.12.1