hBERTv2_new_pretrain_rte
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new on the GLUE RTE dataset. It achieves the following results on the evaluation set:
- Loss: 0.6853
- Accuracy: 0.5596
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 |
---|---|---|---|---|
0.7512 | 1.0 | 20 | 0.6929 | 0.4982 |
0.7204 | 2.0 | 40 | 0.6908 | 0.5271 |
0.701 | 3.0 | 60 | 0.6853 | 0.5596 |
0.6315 | 4.0 | 80 | 0.7081 | 0.5632 |
0.5807 | 5.0 | 100 | 0.8746 | 0.5343 |
0.4079 | 6.0 | 120 | 0.8831 | 0.5632 |
0.3077 | 7.0 | 140 | 1.0779 | 0.5487 |
0.2453 | 8.0 | 160 | 1.1810 | 0.5415 |
Framework versions
- Transformers 4.29.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 4