README.md

metadata

license: gpl-3.0
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: bert-base-chinese-finetuned-ner_0301_J_DATA
    results: []

bert-base-chinese-finetuned-ner_0301_J_DATA

This model is a fine-tuned version of ckiplab/bert-base-chinese-ner on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0318
• Precision: 0.9551
• Recall: 0.9787
• F1: 0.9668
• Accuracy: 0.9923

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.352	1.0	705	0.0754	0.8558	0.9182	0.8859	0.9774
0.0636	2.0	1410	0.0928	0.9082	0.9428	0.9252	0.9794
0.025	3.0	2115	0.0576	0.9262	0.9574	0.9416	0.9828
0.0253	4.0	2820	0.0801	0.9419	0.9630	0.9523	0.9824
0.0169	5.0	3525	0.0400	0.9287	0.9641	0.9461	0.9886
0.0108	6.0	4230	0.0370	0.9372	0.9709	0.9537	0.9903
0.0143	7.0	4935	0.0430	0.9308	0.9652	0.9477	0.9855
0.0083	8.0	5640	0.0648	0.9382	0.9709	0.9543	0.9877
0.0057	9.0	6345	0.0269	0.9222	0.9697	0.9454	0.9903
0.0036	10.0	7050	0.0338	0.9464	0.9697	0.9579	0.9927
0.003	11.0	7755	0.0486	0.9581	0.9742	0.9661	0.9894
0.0017	12.0	8460	0.0230	0.9593	0.9765	0.9678	0.9909
0.001	13.0	9165	0.0260	0.9508	0.9753	0.9629	0.9949
0.0014	14.0	9870	0.0357	0.9582	0.9765	0.9672	0.9914
0.0008	15.0	10575	0.0318	0.9551	0.9787	0.9668	0.9923

Framework versions

• Transformers 4.20.1
• Pytorch 1.13.0+cu117
• Datasets 2.8.0
• Tokenizers 0.12.1