BERTreach-finetuned-ner
This model is a fine-tuned version of jimregan/BERTreach on the wikiann dataset. It achieves the following results on the evaluation set:
- Loss: 0.4944
- Precision: 0.5201
- Recall: 0.5667
- F1: 0.5424
- Accuracy: 0.8366
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: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 63 | 0.7249 | 0.3645 | 0.3905 | 0.3770 | 0.7584 |
No log | 2.0 | 126 | 0.5850 | 0.4529 | 0.4948 | 0.4729 | 0.8072 |
No log | 3.0 | 189 | 0.5192 | 0.4949 | 0.5456 | 0.5190 | 0.8288 |
No log | 4.0 | 252 | 0.5042 | 0.5208 | 0.5592 | 0.5393 | 0.8348 |
No log | 5.0 | 315 | 0.4944 | 0.5201 | 0.5667 | 0.5424 | 0.8366 |
Framework versions
- Transformers 4.12.5
- Pytorch 1.10.0+cu111
- Datasets 1.16.1
- Tokenizers 0.10.3
- Downloads last month
- 6
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Dataset used to train jimregan/BERTreach-finetuned-ner
Evaluation results
- Precision on wikiannself-reported0.520
- Recall on wikiannself-reported0.567
- F1 on wikiannself-reported0.542
- Accuracy on wikiannself-reported0.837