bert-finetuned-ner
This model is a fine-tuned version of FacebookAI/roberta-large on the few-nerd dataset. It achieves the following results on the evaluation set:
- Loss: 0.2164
- Precision: 0.7845
- Recall: 0.8148
- F1: 0.7993
- Accuracy: 0.9429
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1953 | 1.0 | 32942 | 0.1933 | 0.7670 | 0.7968 | 0.7816 | 0.9395 |
| 0.1573 | 2.0 | 65884 | 0.2051 | 0.7850 | 0.8034 | 0.7941 | 0.9416 |
| 0.1256 | 3.0 | 98826 | 0.2164 | 0.7845 | 0.8148 | 0.7993 | 0.9429 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.1+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 38
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for mdroth/bert-finetuned-ner
Base model
FacebookAI/roberta-largeEvaluation results
- Precision on few-nerdvalidation set self-reported0.784
- Recall on few-nerdvalidation set self-reported0.815
- F1 on few-nerdvalidation set self-reported0.799
- Accuracy on few-nerdvalidation set self-reported0.943