<!-- 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_logit_kd_pretrain_mrpc
This model is a fine-tuned version of gokuls/distilbert_sa_pre-training-complete on the GLUE MRPC dataset. It achieves the following results on the evaluation set:
- Loss: 0.4973
- Accuracy: 0.5270
- F1: 0.5089
- Combined Score: 0.5179
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 |
---|---|---|---|---|---|---|
0.5301 | 1.0 | 15 | 0.5200 | 0.3162 | 0.0 | 0.1581 |
0.5164 | 2.0 | 30 | 0.5159 | 0.3162 | 0.0 | 0.1581 |
0.4983 | 3.0 | 45 | 0.5141 | 0.4657 | 0.4076 | 0.4366 |
0.4715 | 4.0 | 60 | 0.5215 | 0.6642 | 0.7187 | 0.6915 |
0.4454 | 5.0 | 75 | 0.4973 | 0.5270 | 0.5089 | 0.5179 |
0.4184 | 6.0 | 90 | 0.4984 | 0.5711 | 0.5803 | 0.5757 |
0.4027 | 7.0 | 105 | 0.5033 | 0.6152 | 0.6424 | 0.6288 |
0.399 | 8.0 | 120 | 0.5095 | 0.7230 | 0.7797 | 0.7514 |
0.3975 | 9.0 | 135 | 0.5087 | 0.6176 | 0.6486 | 0.6331 |
0.3979 | 10.0 | 150 | 0.5088 | 0.7010 | 0.7579 | 0.7295 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2