metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
model-index:
- name: model
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: glue
type: glue
args: sst2
metrics:
- name: Accuracy
type: accuracy
value: 0.8302752293577982
model
This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on the glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.9810
- Accuracy: 0.8303
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: 0.00010445576414788915
- train_batch_size: 1024
- eval_batch_size: 1024
- seed: 33
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.6492 | 1.0 | 66 | 1.2487 | 0.7844 |
| 1.0585 | 2.0 | 132 | 1.0561 | 0.8073 |
| 0.8081 | 3.0 | 198 | 0.9585 | 0.8154 |
| 0.6595 | 4.0 | 264 | 0.9454 | 0.8268 |
| 0.5681 | 5.0 | 330 | 0.9372 | 0.8257 |
| 0.512 | 6.0 | 396 | 0.9471 | 0.8303 |
| 0.4868 | 7.0 | 462 | 0.9803 | 0.8291 |
| 0.4643 | 8.0 | 528 | 0.9699 | 0.8326 |
| 0.4498 | 9.0 | 594 | 0.9791 | 0.8280 |
| 0.4402 | 10.0 | 660 | 0.9810 | 0.8303 |
Framework versions
- Transformers 4.12.3
- Pytorch 1.9.1
- Datasets 1.15.1
- Tokenizers 0.10.3