<!-- 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-large-da-multiwoz2.0_400-ep20-nonstop
This model is a fine-tuned version of google/flan-t5-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3661
- Accuracy: 41.2421
- Num: 7358
- Gen Len: 15.5222
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 64
- seed: 1799
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Num | Gen Len |
---|---|---|---|---|---|---|
1.1824 | 1.16 | 200 | 0.5187 | 28.4524 | 7358 | 14.7642 |
0.5471 | 2.33 | 400 | 0.4278 | 32.5629 | 7358 | 15.4386 |
0.4647 | 3.49 | 600 | 0.4029 | 35.2443 | 7358 | 16.135 |
0.4313 | 4.65 | 800 | 0.3820 | 36.6479 | 7358 | 16.1552 |
0.4074 | 5.81 | 1000 | 0.3775 | 37.6957 | 7358 | 15.1439 |
0.3859 | 6.98 | 1200 | 0.3690 | 38.3142 | 7358 | 15.2045 |
0.369 | 8.14 | 1400 | 0.3720 | 39.8799 | 7358 | 15.7923 |
0.3547 | 9.3 | 1600 | 0.3665 | 39.5217 | 7358 | 15.3394 |
0.3457 | 10.47 | 1800 | 0.3632 | 39.8289 | 7358 | 15.4761 |
0.3423 | 11.63 | 2000 | 0.3678 | 39.9509 | 7358 | 15.6708 |
0.3295 | 12.79 | 2200 | 0.3657 | 41.1373 | 7358 | 15.1586 |
0.3212 | 13.95 | 2400 | 0.3651 | 40.8611 | 7358 | 15.7312 |
0.3128 | 15.12 | 2600 | 0.3664 | 40.8806 | 7358 | 15.4553 |
0.3131 | 16.28 | 2800 | 0.3677 | 40.8906 | 7358 | 15.4629 |
0.3093 | 17.44 | 3000 | 0.3661 | 40.9971 | 7358 | 15.4329 |
0.3021 | 18.6 | 3200 | 0.3652 | 41.2953 | 7358 | 15.5118 |
0.3004 | 19.77 | 3400 | 0.3661 | 41.2492 | 7358 | 15.5246 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.0+cu111
- Datasets 2.5.1
- Tokenizers 0.12.1