hBERTv2_new_pretrain_w_init__stsb
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new_wt_init on the GLUE STSB dataset. It achieves the following results on the evaluation set:
- Loss: 2.0270
- Pearson: 0.3743
- Spearmanr: 0.3670
- Combined Score: 0.3707
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 | Pearson | Spearmanr | Combined Score |
---|---|---|---|---|---|---|
2.2654 | 1.0 | 45 | 2.4836 | 0.2041 | 0.1912 | 0.1976 |
1.9657 | 2.0 | 90 | 2.1138 | 0.2744 | 0.2547 | 0.2646 |
1.6665 | 3.0 | 135 | 2.2375 | 0.3087 | 0.3002 | 0.3044 |
1.3265 | 4.0 | 180 | 2.0270 | 0.3743 | 0.3670 | 0.3707 |
1.0731 | 5.0 | 225 | 2.3748 | 0.3294 | 0.3212 | 0.3253 |
0.7974 | 6.0 | 270 | 2.6753 | 0.3338 | 0.3353 | 0.3345 |
0.6738 | 7.0 | 315 | 2.5125 | 0.3590 | 0.3464 | 0.3527 |
0.5384 | 8.0 | 360 | 2.3740 | 0.3310 | 0.3211 | 0.3261 |
0.4589 | 9.0 | 405 | 2.3911 | 0.3709 | 0.3690 | 0.3699 |
Framework versions
- Transformers 4.29.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 7
Dataset used to train gokuls/hBERTv2_new_pretrain_w_init__stsb
Evaluation results
- Spearmanr on GLUE STSBvalidation set self-reported0.367