metadata
license: apache-2.0
base_model: bert-base-uncased
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- matthews_correlation
model-index:
- name: bert-base-uncased-cola-lora-epochs-10-lr-0.0005
results: []
bert-base-uncased-cola-lora-epochs-10-lr-0.0005
This model is a fine-tuned version of bert-base-uncased on the glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.7759
- Matthews Correlation: 0.6347
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: 0.0005
- train_batch_size: 32
- eval_batch_size: 32
- seed: 28
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|---|---|---|---|---|
| No log | 1.0 | 265 | 0.6135 | 0.2906 |
| 0.4571 | 2.0 | 530 | 0.4470 | 0.4851 |
| 0.4571 | 3.0 | 795 | 0.3926 | 0.5648 |
| 0.2553 | 4.0 | 1060 | 0.4631 | 0.5358 |
| 0.2553 | 5.0 | 1325 | 0.4592 | 0.5145 |
| 0.1378 | 6.0 | 1590 | 0.8133 | 0.5608 |
| 0.1378 | 7.0 | 1855 | 0.6745 | 0.6815 |
| 0.081 | 8.0 | 2120 | 0.6891 | 0.6815 |
| 0.081 | 9.0 | 2385 | 0.6823 | 0.6110 |
| 0.0512 | 10.0 | 2650 | 0.7759 | 0.6347 |
Framework versions
- Transformers 4.32.0.dev0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3