hBERTv1_new_pretrain_48_KD_mnli
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_48_KD on the GLUE MNLI dataset. It achieves the following results on the evaluation set:
- Loss: 1.0977
- Accuracy: 0.3268
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: 4e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.1011 | 1.0 | 3068 | 1.1001 | 0.3176 |
| 1.0991 | 2.0 | 6136 | 1.0978 | 0.3295 |
| 1.0983 | 3.0 | 9204 | 1.0985 | 0.3351 |
| 1.0984 | 4.0 | 12272 | 1.0983 | 0.3294 |
| 1.098 | 5.0 | 15340 | 1.0978 | 0.3264 |
| 1.098 | 6.0 | 18408 | 1.0979 | 0.3285 |
| 1.098 | 7.0 | 21476 | 1.0980 | 0.3272 |
| 1.098 | 8.0 | 24544 | 1.0981 | 0.3266 |
| 1.098 | 9.0 | 27612 | 1.0980 | 0.3256 |
| 1.098 | 10.0 | 30680 | 1.0985 | 0.3320 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3