hBERTv1_new_pretrain_48_emb_com_qnli
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48 on the GLUE QNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.6611
- Accuracy: 0.5984
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 |
|---|---|---|---|---|
| 0.6862 | 1.0 | 819 | 0.6995 | 0.5513 |
| 0.6677 | 2.0 | 1638 | 0.6624 | 0.5949 |
| 0.6553 | 3.0 | 2457 | 0.6611 | 0.5984 |
| 0.6426 | 4.0 | 3276 | 0.6866 | 0.5762 |
| 0.6244 | 5.0 | 4095 | 0.6751 | 0.5951 |
| 0.6136 | 6.0 | 4914 | 0.6903 | 0.5728 |
| 0.6036 | 7.0 | 5733 | 0.7188 | 0.5731 |
| 0.5855 | 8.0 | 6552 | 0.7175 | 0.5825 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3