--- tags: - generated_from_trainer metrics: - precision - recall - f1 - accuracy model-index: - name: bert-base-chinese-david-ner results: [] --- # bert-base-chinese-david-ner This model is a fine-tuned version of [bert-base-chinese](https://huggingface.co/bert-base-chinese) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.0530 - Precision: 0.9486 - Recall: 0.9619 - F1: 0.9552 - Accuracy: 0.9892 ## 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: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 4 ### Training results | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| | 0.9951 | 0.1 | 100 | 0.3761 | 0.4443 | 0.4460 | 0.4451 | 0.8659 | | 0.2378 | 0.2 | 200 | 0.1500 | 0.7185 | 0.7980 | 0.7562 | 0.9512 | | 0.1625 | 0.3 | 300 | 0.1324 | 0.7662 | 0.8247 | 0.7944 | 0.9546 | | 0.1351 | 0.39 | 400 | 0.1014 | 0.8024 | 0.8615 | 0.8309 | 0.9665 | | 0.1083 | 0.49 | 500 | 0.0814 | 0.8529 | 0.9136 | 0.8822 | 0.9771 | | 0.0944 | 0.59 | 600 | 0.1071 | 0.8282 | 0.8945 | 0.8601 | 0.9683 | | 0.106 | 0.69 | 700 | 0.0696 | 0.8784 | 0.9085 | 0.8932 | 0.9802 | | 0.0897 | 0.79 | 800 | 0.0743 | 0.8764 | 0.9098 | 0.8928 | 0.9790 | | 0.0809 | 0.89 | 900 | 0.0677 | 0.8933 | 0.9250 | 0.9089 | 0.9813 | | 0.0856 | 0.99 | 1000 | 0.0646 | 0.8966 | 0.9250 | 0.9106 | 0.9808 | | 0.057 | 1.09 | 1100 | 0.0719 | 0.8851 | 0.9199 | 0.9022 | 0.9800 | | 0.0592 | 1.18 | 1200 | 0.0670 | 0.9064 | 0.9225 | 0.9144 | 0.9816 | | 0.0461 | 1.28 | 1300 | 0.0593 | 0.9081 | 0.9288 | 0.9183 | 0.9836 | | 0.0555 | 1.38 | 1400 | 0.0576 | 0.9151 | 0.9314 | 0.9232 | 0.9848 | | 0.0582 | 1.48 | 1500 | 0.0592 | 0.9167 | 0.9365 | 0.9265 | 0.9854 | | 0.0511 | 1.58 | 1600 | 0.0538 | 0.9203 | 0.9390 | 0.9296 | 0.9857 | | 0.0438 | 1.68 | 1700 | 0.0568 | 0.9081 | 0.9288 | 0.9183 | 0.9848 | | 0.0486 | 1.78 | 1800 | 0.0546 | 0.9160 | 0.9428 | 0.9292 | 0.9859 | | 0.0534 | 1.88 | 1900 | 0.0472 | 0.9232 | 0.9466 | 0.9348 | 0.9863 | | 0.0432 | 1.97 | 2000 | 0.0457 | 0.9283 | 0.9543 | 0.9411 | 0.9877 | | 0.0355 | 2.07 | 2100 | 0.0530 | 0.9376 | 0.9543 | 0.9458 | 0.9882 | | 0.0234 | 2.17 | 2200 | 0.0595 | 0.9327 | 0.9504 | 0.9415 | 0.9867 | | 0.0306 | 2.27 | 2300 | 0.0485 | 0.9377 | 0.9568 | 0.9472 | 0.9891 | | 0.0317 | 2.37 | 2400 | 0.0464 | 0.9437 | 0.9593 | 0.9515 | 0.9897 | | 0.0234 | 2.47 | 2500 | 0.0510 | 0.9410 | 0.9530 | 0.9470 | 0.9884 | | 0.0276 | 2.57 | 2600 | 0.0507 | 0.9342 | 0.9568 | 0.9454 | 0.9883 | | 0.0279 | 2.67 | 2700 | 0.0495 | 0.9384 | 0.9492 | 0.9438 | 0.9875 | | 0.0292 | 2.76 | 2800 | 0.0491 | 0.9388 | 0.9555 | 0.9471 | 0.9880 | | 0.0209 | 2.86 | 2900 | 0.0502 | 0.9451 | 0.9619 | 0.9534 | 0.9879 | | 0.0257 | 2.96 | 3000 | 0.0510 | 0.9438 | 0.9606 | 0.9521 | 0.9878 | | 0.0205 | 3.06 | 3100 | 0.0512 | 0.9391 | 0.9593 | 0.9491 | 0.9884 | | 0.0112 | 3.16 | 3200 | 0.0529 | 0.9391 | 0.9606 | 0.9497 | 0.9887 | | 0.017 | 3.26 | 3300 | 0.0614 | 0.9414 | 0.9593 | 0.9503 | 0.9874 | | 0.0178 | 3.36 | 3400 | 0.0490 | 0.9353 | 0.9555 | 0.9453 | 0.9889 | | 0.0189 | 3.46 | 3500 | 0.0512 | 0.9426 | 0.9606 | 0.9515 | 0.9889 | | 0.0175 | 3.55 | 3600 | 0.0539 | 0.9426 | 0.9593 | 0.9509 | 0.9884 | | 0.012 | 3.65 | 3700 | 0.0521 | 0.945 | 0.9606 | 0.9527 | 0.9894 | | 0.0179 | 3.75 | 3800 | 0.0533 | 0.9438 | 0.9606 | 0.9521 | 0.9886 | | 0.0185 | 3.85 | 3900 | 0.0530 | 0.9486 | 0.9619 | 0.9552 | 0.9892 | | 0.0175 | 3.95 | 4000 | 0.0522 | 0.9462 | 0.9606 | 0.9533 | 0.9890 | ### Framework versions - Transformers 4.29.0.dev0 - Pytorch 1.10.1+cu113 - Datasets 2.11.0 - Tokenizers 0.13.3