funnel-transformer-xlarge_ner_wnut_17
This model is a fine-tuned version of funnel-transformer/xlarge on the wnut_17 dataset. It achieves the following results on the evaluation set:
- Loss: 0.2453
- Precision: 0.7205
- Recall: 0.5921
- F1: 0.6500
- Accuracy: 0.9620
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: cosine
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 213 | 0.2331 | 0.6897 | 0.4067 | 0.5117 | 0.9462 |
| No log | 2.0 | 426 | 0.2056 | 0.7097 | 0.5526 | 0.6214 | 0.9587 |
| 0.1454 | 3.0 | 639 | 0.2379 | 0.7102 | 0.5658 | 0.6298 | 0.9600 |
| 0.1454 | 4.0 | 852 | 0.2397 | 0.7141 | 0.5885 | 0.6452 | 0.9620 |
| 0.0319 | 5.0 | 1065 | 0.2453 | 0.7205 | 0.5921 | 0.6500 | 0.9620 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1