<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
distilbert_sa_GLUE_Experiment_qqp
This model is a fine-tuned version of distilbert-base-uncased on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.4299
- Accuracy: 0.7981
- F1: 0.7243
- Combined Score: 0.7612
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 | F1 | Combined Score |
---|---|---|---|---|---|---|
0.5166 | 1.0 | 1422 | 0.4817 | 0.7654 | 0.6882 | 0.7268 |
0.4462 | 2.0 | 2844 | 0.4460 | 0.7885 | 0.6950 | 0.7417 |
0.3979 | 3.0 | 4266 | 0.4299 | 0.7981 | 0.7243 | 0.7612 |
0.3497 | 4.0 | 5688 | 0.4417 | 0.7972 | 0.7421 | 0.7696 |
0.2994 | 5.0 | 7110 | 0.4330 | 0.8099 | 0.7495 | 0.7797 |
0.2514 | 6.0 | 8532 | 0.4764 | 0.8137 | 0.7499 | 0.7818 |
0.2065 | 7.0 | 9954 | 0.4819 | 0.8123 | 0.7520 | 0.7822 |
0.1669 | 8.0 | 11376 | 0.5460 | 0.8132 | 0.7555 | 0.7844 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2