<!-- 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. -->
FLAN-T5_GLUE_finetuning_lr2e-4
This model is a fine-tuned version of google/flan-t5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0894
- Accuracy: 0.8851
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: 0.0002
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.1301 | 0.17 | 2500 | 0.1079 | 0.8332 |
0.1061 | 0.34 | 5000 | 0.0960 | 0.8507 |
0.1002 | 0.51 | 7500 | 0.0910 | 0.8577 |
0.0969 | 0.68 | 10000 | 0.0881 | 0.8595 |
0.0934 | 0.85 | 12500 | 0.0858 | 0.8659 |
0.0912 | 1.02 | 15000 | 0.0846 | 0.8672 |
0.076 | 1.18 | 17500 | 0.0828 | 0.8713 |
0.0765 | 1.35 | 20000 | 0.0840 | 0.8715 |
0.0764 | 1.52 | 22500 | 0.0818 | 0.872 |
0.0757 | 1.69 | 25000 | 0.0823 | 0.8746 |
0.076 | 1.86 | 27500 | 0.0811 | 0.8747 |
0.0723 | 2.03 | 30000 | 0.0836 | 0.8744 |
0.0593 | 2.2 | 32500 | 0.0822 | 0.8773 |
0.0603 | 2.37 | 35000 | 0.0843 | 0.8769 |
0.0599 | 2.54 | 37500 | 0.0823 | 0.8784 |
0.0608 | 2.71 | 40000 | 0.0808 | 0.879 |
0.06 | 2.88 | 42500 | 0.0797 | 0.8809 |
0.0564 | 3.05 | 45000 | 0.0872 | 0.881 |
0.0476 | 3.22 | 47500 | 0.0871 | 0.8816 |
0.0473 | 3.38 | 50000 | 0.0866 | 0.8811 |
0.0474 | 3.55 | 52500 | 0.0851 | 0.8837 |
0.0485 | 3.72 | 55000 | 0.0828 | 0.8851 |
0.048 | 3.89 | 57500 | 0.0837 | 0.885 |
0.0445 | 4.06 | 60000 | 0.0887 | 0.884 |
0.0383 | 4.23 | 62500 | 0.0917 | 0.8853 |
0.0392 | 4.4 | 65000 | 0.0901 | 0.8843 |
0.0393 | 4.57 | 67500 | 0.0896 | 0.8848 |
0.039 | 4.74 | 70000 | 0.0896 | 0.8851 |
0.0389 | 4.91 | 72500 | 0.0894 | 0.8851 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.0+cu117
- Datasets 2.10.1
- Tokenizers 0.12.1