<!-- 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_cola
This model is a fine-tuned version of gokuls/distilbert_sa_pre-training-complete on the GLUE COLA dataset. It achieves the following results on the evaluation set:
- Loss: 0.4135
- Matthews Correlation: 0.5181
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 | Matthews Correlation |
---|---|---|---|---|
0.7453 | 1.0 | 34 | 0.5042 | 0.3040 |
0.4991 | 2.0 | 68 | 0.4737 | 0.3891 |
0.3299 | 3.0 | 102 | 0.4457 | 0.4447 |
0.2426 | 4.0 | 136 | 0.4319 | 0.5037 |
0.1889 | 5.0 | 170 | 0.4813 | 0.4461 |
0.1552 | 6.0 | 204 | 0.4379 | 0.4992 |
0.1412 | 7.0 | 238 | 0.4435 | 0.4943 |
0.1263 | 8.0 | 272 | 0.4172 | 0.5272 |
0.114 | 9.0 | 306 | 0.5057 | 0.4884 |
0.1086 | 10.0 | 340 | 0.4487 | 0.5265 |
0.0981 | 11.0 | 374 | 0.4135 | 0.5181 |
0.0949 | 12.0 | 408 | 0.4759 | 0.4885 |
0.0882 | 13.0 | 442 | 0.4691 | 0.4747 |
0.0904 | 14.0 | 476 | 0.4164 | 0.5040 |
0.0847 | 15.0 | 510 | 0.5206 | 0.4692 |
0.0828 | 16.0 | 544 | 0.4488 | 0.4874 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2