hBERTv1_new_pretrain_qqp
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.4260
- Accuracy: 0.8042
- F1: 0.7096
- Combined Score: 0.7569
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 | F1 | Combined Score |
---|---|---|---|---|---|---|
0.5025 | 1.0 | 2843 | 0.4596 | 0.7833 | 0.6804 | 0.7318 |
0.4098 | 2.0 | 5686 | 0.4260 | 0.8042 | 0.7096 | 0.7569 |
0.3558 | 3.0 | 8529 | 0.4301 | 0.8095 | 0.7216 | 0.7656 |
0.3086 | 4.0 | 11372 | 0.4599 | 0.8174 | 0.7536 | 0.7855 |
0.2718 | 5.0 | 14215 | 0.4817 | 0.8203 | 0.7447 | 0.7825 |
0.2425 | 6.0 | 17058 | 0.4866 | 0.8220 | 0.7569 | 0.7895 |
0.2161 | 7.0 | 19901 | 0.4497 | 0.8251 | 0.7608 | 0.7929 |
Framework versions
- Transformers 4.29.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 8
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Dataset used to train gokuls/hBERTv1_new_pretrain_qqp
Evaluation results
- Accuracy on GLUE QQPvalidation set self-reported0.804
- F1 on GLUE QQPvalidation set self-reported0.710