distilbert_add_GLUE_Experiment_logit_kd_pretrain_rte
This model is a fine-tuned version of gokuls/distilbert_add_pre-training-complete on the GLUE RTE dataset. It achieves the following results on the evaluation set:
- Loss: 0.4234
- Accuracy: 0.4729
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
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.4287 | 1.0 | 10 | 0.4234 | 0.4729 |
| 0.417 | 2.0 | 20 | 0.4299 | 0.4729 |
| 0.416 | 3.0 | 30 | 0.4256 | 0.4729 |
| 0.4155 | 4.0 | 40 | 0.4249 | 0.4765 |
| 0.4121 | 5.0 | 50 | 0.4279 | 0.4801 |
| 0.4089 | 6.0 | 60 | 0.4273 | 0.4765 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2