<!-- 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. -->
bert-base-cased-conversational-ner
This model is a fine-tuned version of DeepPavlov/bert-base-cased-conversational on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3583
- Job Title precision: 0.8377
- Job Title recall: 0.8317
- Job Title f1: 0.8347
- Loc precision: 0.8938
- Loc recall: 0.9340
- Loc f1: 0.9135
- Org precision: 0.7092
- Org recall: 0.7032
- Org f1: 0.7062
- Misc precision: 0.6246
- Misc recall: 0.7270
- Misc f1: 0.6719
- Precision: 0.8154
- Recall: 0.8240
- F1: 0.8197
- Accuracy: 0.8687
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
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Job Title precision | Job Title recall | Job Title f1 | Loc precision | Loc recall | Loc f1 | Org precision | Org recall | Org f1 | Misc precision | Misc recall | Misc f1 | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
No log | 1.0 | 308 | 0.3583 | 0.8377 | 0.8317 | 0.8347 | 0.8938 | 0.9340 | 0.9135 | 0.7092 | 0.7032 | 0.7062 | 0.6246 | 0.7270 | 0.6719 | 0.8154 | 0.8240 | 0.8197 | 0.8687 |
0.3975 | 2.0 | 616 | 0.3767 | 0.7906 | 0.9035 | 0.8433 | 0.8731 | 0.9614 | 0.9151 | 0.6275 | 0.7973 | 0.7023 | 0.6623 | 0.6894 | 0.6756 | 0.7658 | 0.8866 | 0.8218 | 0.8669 |
Framework versions
- Transformers 4.28.1
- Pytorch 1.7.1+cu110
- Datasets 2.12.0
- Tokenizers 0.13.2