FLAN-T5_GLUE_finetuning_lr3e-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: 0.0936
- Accuracy: 0.8849
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: 0.0003
- 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 |
---|---|---|---|---|
0.1314 | 0.17 | 2500 | 0.1069 | 0.8312 |
0.1087 | 0.34 | 5000 | 0.0960 | 0.8513 |
0.1029 | 0.51 | 7500 | 0.0913 | 0.8543 |
0.1007 | 0.68 | 10000 | 0.0890 | 0.8587 |
0.0961 | 0.85 | 12500 | 0.0870 | 0.865 |
0.0933 | 1.02 | 15000 | 0.0875 | 0.8644 |
0.0777 | 1.18 | 17500 | 0.0844 | 0.869 |
0.0784 | 1.35 | 20000 | 0.0843 | 0.8701 |
0.078 | 1.52 | 22500 | 0.0827 | 0.8712 |
0.0775 | 1.69 | 25000 | 0.0841 | 0.8711 |
0.0774 | 1.86 | 27500 | 0.0824 | 0.8737 |
0.0732 | 2.03 | 30000 | 0.0862 | 0.8727 |
0.0591 | 2.2 | 32500 | 0.0842 | 0.8766 |
0.06 | 2.37 | 35000 | 0.0858 | 0.8762 |
0.0595 | 2.54 | 37500 | 0.0843 | 0.8781 |
0.061 | 2.71 | 40000 | 0.0818 | 0.88 |
0.0598 | 2.88 | 42500 | 0.0813 | 0.8803 |
0.0558 | 3.05 | 45000 | 0.0896 | 0.8805 |
0.0448 | 3.22 | 47500 | 0.0885 | 0.881 |
0.0451 | 3.38 | 50000 | 0.0888 | 0.8805 |
0.0449 | 3.55 | 52500 | 0.0873 | 0.882 |
0.0457 | 3.72 | 55000 | 0.0853 | 0.8839 |
0.0456 | 3.89 | 57500 | 0.0856 | 0.8835 |
0.0413 | 4.06 | 60000 | 0.0931 | 0.8828 |
0.0342 | 4.23 | 62500 | 0.0961 | 0.8851 |
0.0348 | 4.4 | 65000 | 0.0945 | 0.8851 |
0.0351 | 4.57 | 67500 | 0.0936 | 0.8845 |
0.0346 | 4.74 | 70000 | 0.0940 | 0.8846 |
0.0347 | 4.91 | 72500 | 0.0936 | 0.8849 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.0+cu117
- Datasets 2.10.1
- Tokenizers 0.12.1
- Downloads last month
- 5
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.