<!-- 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_lr1e-3
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: 0.0998
- Accuracy: 0.871
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: 0.001
- 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 |
---|---|---|---|---|
0.1625 | 0.17 | 2500 | 0.1357 | 0.7747 |
0.1352 | 0.34 | 5000 | 0.1165 | 0.8158 |
0.1302 | 0.51 | 7500 | 0.1143 | 0.8193 |
0.1239 | 0.68 | 10000 | 0.1115 | 0.8279 |
0.1175 | 0.85 | 12500 | 0.1050 | 0.8343 |
0.1145 | 1.02 | 15000 | 0.1046 | 0.8339 |
0.1026 | 1.18 | 17500 | 0.0982 | 0.8435 |
0.1016 | 1.35 | 20000 | 0.0981 | 0.8465 |
0.1004 | 1.52 | 22500 | 0.0953 | 0.8495 |
0.0992 | 1.69 | 25000 | 0.0967 | 0.85 |
0.0984 | 1.86 | 27500 | 0.0926 | 0.8542 |
0.0944 | 2.03 | 30000 | 0.0951 | 0.8512 |
0.0813 | 2.2 | 32500 | 0.0933 | 0.8572 |
0.081 | 2.37 | 35000 | 0.0917 | 0.858 |
0.0806 | 2.54 | 37500 | 0.0949 | 0.8608 |
0.0805 | 2.71 | 40000 | 0.0901 | 0.8615 |
0.0786 | 2.88 | 42500 | 0.0889 | 0.8624 |
0.0733 | 3.05 | 45000 | 0.0974 | 0.8625 |
0.0599 | 3.22 | 47500 | 0.0942 | 0.8644 |
0.0596 | 3.38 | 50000 | 0.0950 | 0.8655 |
0.0595 | 3.55 | 52500 | 0.0921 | 0.866 |
0.0602 | 3.72 | 55000 | 0.0912 | 0.8678 |
0.0598 | 3.89 | 57500 | 0.0907 | 0.8684 |
0.053 | 4.06 | 60000 | 0.1013 | 0.8678 |
0.0421 | 4.23 | 62500 | 0.1010 | 0.8697 |
0.0426 | 4.4 | 65000 | 0.1025 | 0.8688 |
0.0427 | 4.57 | 67500 | 0.1023 | 0.8695 |
0.0415 | 4.74 | 70000 | 0.1004 | 0.8711 |
0.0411 | 4.91 | 72500 | 0.0998 | 0.871 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.0+cu117
- Datasets 2.10.1
- Tokenizers 0.12.1