metadata
library_name: transformers
language:
- en
license: apache-2.0
base_model: google/bert_uncased_L-4_H-256_A-4
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- matthews_correlation
- accuracy
model-index:
- name: bert_uncased_L-4_H-256_A-4_cola
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE COLA
type: glue
args: cola
metrics:
- name: Matthews Correlation
type: matthews_correlation
value: 0.2650812590803394
- name: Accuracy
type: accuracy
value: 0.7027804255485535
bert_uncased_L-4_H-256_A-4_cola
This model is a fine-tuned version of google/bert_uncased_L-4_H-256_A-4 on the GLUE COLA dataset. It achieves the following results on the evaluation set:
- Loss: 0.5943
- Matthews Correlation: 0.2651
- Accuracy: 0.7028
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: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation | Accuracy |
|---|---|---|---|---|---|
| 0.6358 | 1.0 | 34 | 0.6182 | 0.0 | 0.6913 |
| 0.6077 | 2.0 | 68 | 0.6184 | 0.0 | 0.6913 |
| 0.5982 | 3.0 | 102 | 0.6035 | 0.0 | 0.6913 |
| 0.575 | 4.0 | 136 | 0.5997 | 0.1458 | 0.7009 |
| 0.5391 | 5.0 | 170 | 0.5992 | 0.2018 | 0.7028 |
| 0.4999 | 6.0 | 204 | 0.6159 | 0.2088 | 0.7085 |
| 0.4722 | 7.0 | 238 | 0.5974 | 0.2782 | 0.7248 |
| 0.4437 | 8.0 | 272 | 0.5943 | 0.2651 | 0.7028 |
| 0.4204 | 9.0 | 306 | 0.6239 | 0.2618 | 0.7210 |
| 0.3956 | 10.0 | 340 | 0.6360 | 0.2655 | 0.7191 |
| 0.3671 | 11.0 | 374 | 0.6876 | 0.2592 | 0.7200 |
| 0.3546 | 12.0 | 408 | 0.7041 | 0.2665 | 0.7239 |
| 0.333 | 13.0 | 442 | 0.6849 | 0.2891 | 0.7229 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.2.1+cu118
- Datasets 2.17.0
- Tokenizers 0.20.3