<!-- 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_mnli
This model is a fine-tuned version of distilbert-base-uncased on the GLUE MNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.4989
- Accuracy: 0.6525
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 |
---|---|---|---|---|
0.575 | 1.0 | 1534 | 0.5428 | 0.5554 |
0.5345 | 2.0 | 3068 | 0.5205 | 0.5987 |
0.511 | 3.0 | 4602 | 0.5105 | 0.6222 |
0.4917 | 4.0 | 6136 | 0.5021 | 0.6360 |
0.4735 | 5.0 | 7670 | 0.5004 | 0.6470 |
0.4557 | 6.0 | 9204 | 0.4976 | 0.6534 |
0.4391 | 7.0 | 10738 | 0.4982 | 0.6606 |
0.4231 | 8.0 | 12272 | 0.4982 | 0.6586 |
0.4082 | 9.0 | 13806 | 0.5020 | 0.6587 |
0.394 | 10.0 | 15340 | 0.5082 | 0.6561 |
0.3816 | 11.0 | 16874 | 0.5140 | 0.6617 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2