metadata
license: mit
base_model: surrey-nlp/roberta-large-finetuned-abbr
tags:
- generated_from_trainer
datasets:
- plod-filtered
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: roberta-large-finetuned-abbr-finetuned-ner
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: plod-filtered
type: plod-filtered
config: PLODfiltered
split: validation
args: PLODfiltered
metrics:
- name: Precision
type: precision
value: 0.9800350338833268
- name: Recall
type: recall
value: 0.9766733969309696
- name: F1
type: f1
value: 0.9783513277508114
- name: Accuracy
type: accuracy
value: 0.9761728475392376
roberta-large-finetuned-abbr-finetuned-ner
This model is a fine-tuned version of surrey-nlp/roberta-large-finetuned-abbr on the plod-filtered dataset. It achieves the following results on the evaluation set:
- Loss: 0.0913
- Precision: 0.9800
- Recall: 0.9767
- F1: 0.9784
- Accuracy: 0.9762
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: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0805 | 0.99 | 7000 | 0.0761 | 0.9762 | 0.9722 | 0.9742 | 0.9720 |
| 0.0655 | 1.99 | 14000 | 0.0682 | 0.9769 | 0.9748 | 0.9759 | 0.9735 |
| 0.0469 | 2.98 | 21000 | 0.0718 | 0.9787 | 0.9746 | 0.9767 | 0.9744 |
| 0.0336 | 3.98 | 28000 | 0.0851 | 0.9800 | 0.9753 | 0.9776 | 0.9753 |
| 0.0259 | 4.97 | 35000 | 0.0913 | 0.9800 | 0.9767 | 0.9784 | 0.9762 |
| 0.0197 | 5.97 | 42000 | 0.0948 | 0.9801 | 0.9774 | 0.9787 | 0.9766 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.1+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0