<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
kt_punc
This model is a fine-tuned version of bert-base-chinese on the chn_senti_corp dataset. It achieves the following results on the evaluation set:
- Loss: 0.1703
- Precision: 0.7079
- Recall: 0.7314
- F1: 0.7194
- Accuracy: 0.9573
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.1661 | 1.0 | 600 | 0.1351 | 0.6566 | 0.6833 | 0.6697 | 0.9498 |
0.1246 | 2.0 | 1200 | 0.1330 | 0.6854 | 0.6665 | 0.6758 | 0.9521 |
0.1121 | 3.0 | 1800 | 0.1303 | 0.6885 | 0.6994 | 0.6939 | 0.9537 |
0.1008 | 4.0 | 2400 | 0.1359 | 0.6836 | 0.7248 | 0.7036 | 0.9543 |
0.0809 | 5.0 | 3000 | 0.1404 | 0.7035 | 0.7082 | 0.7059 | 0.9559 |
0.0696 | 6.0 | 3600 | 0.1449 | 0.6986 | 0.7224 | 0.7103 | 0.9560 |
0.0628 | 7.0 | 4200 | 0.1563 | 0.7063 | 0.7214 | 0.7138 | 0.9567 |
0.0561 | 8.0 | 4800 | 0.1618 | 0.7024 | 0.7333 | 0.7175 | 0.9568 |
0.0525 | 9.0 | 5400 | 0.1669 | 0.7083 | 0.7335 | 0.7207 | 0.9574 |
0.0453 | 10.0 | 6000 | 0.1703 | 0.7079 | 0.7314 | 0.7194 | 0.9573 |
Framework versions
- Transformers 4.19.1
- Pytorch 1.11.0+cu113
- Datasets 2.2.1
- Tokenizers 0.12.1