bert_uncased_L-4_H-512_A-8_cola
This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the GLUE COLA dataset. It achieves the following results on the evaluation set:
- Loss: 0.5796
- Matthews Correlation: 0.2588
- Accuracy: 0.7248
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 |
|---|---|---|---|---|---|
| 0.6069 | 1.0 | 34 | 0.6018 | 0.0 | 0.6913 |
| 0.5619 | 2.0 | 68 | 0.5891 | 0.1766 | 0.7076 |
| 0.4858 | 3.0 | 102 | 0.5796 | 0.2588 | 0.7248 |
| 0.4109 | 4.0 | 136 | 0.6467 | 0.2838 | 0.7306 |
| 0.349 | 5.0 | 170 | 0.6379 | 0.3133 | 0.7354 |
| 0.294 | 6.0 | 204 | 0.6805 | 0.3436 | 0.7440 |
| 0.2564 | 7.0 | 238 | 0.7498 | 0.3178 | 0.7363 |
| 0.2222 | 8.0 | 272 | 0.7861 | 0.3320 | 0.7383 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.2.1+cu118
- Datasets 2.17.0
- Tokenizers 0.20.3
- Downloads last month
- 104
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for gokulsrinivasagan/bert_uncased_L-4_H-512_A-8_cola
Base model
google/bert_uncased_L-4_H-512_A-8Dataset used to train gokulsrinivasagan/bert_uncased_L-4_H-512_A-8_cola
Evaluation results
- Matthews Correlation on GLUE COLAself-reported0.259
- Accuracy on GLUE COLAself-reported0.725