<!-- 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-multiwoz2.1_800-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: 0.3425
- Accuracy: 43.5925
- Num: 7365
- Gen Len: 15.6426
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 |
|---|---|---|---|---|---|---|
| 0.8715 | 1.19 | 400 | 0.4399 | 33.0885 | 7365 | 16.6641 |
| 0.4615 | 2.38 | 800 | 0.3853 | 35.0673 | 7365 | 14.4206 |
| 0.4139 | 3.57 | 1200 | 0.3576 | 37.9691 | 7365 | 16.1995 |
| 0.3802 | 4.76 | 1600 | 0.3539 | 39.7252 | 7365 | 15.447 |
| 0.3643 | 5.95 | 2000 | 0.3436 | 40.0614 | 7365 | 15.7359 |
| 0.347 | 7.14 | 2400 | 0.3438 | 41.028 | 7365 | 15.2403 |
| 0.3351 | 8.33 | 2800 | 0.3392 | 42.0823 | 7365 | 15.7405 |
| 0.3269 | 9.52 | 3200 | 0.3372 | 42.2902 | 7365 | 15.7686 |
| 0.3165 | 10.71 | 3600 | 0.3381 | 42.3669 | 7365 | 15.4957 |
| 0.3093 | 11.9 | 4000 | 0.3405 | 42.9291 | 7365 | 15.6539 |
| 0.3018 | 13.1 | 4400 | 0.3421 | 42.9564 | 7365 | 15.8058 |
| 0.2962 | 14.29 | 4800 | 0.3381 | 43.2625 | 7365 | 15.5039 |
| 0.2893 | 15.48 | 5200 | 0.3424 | 43.5069 | 7365 | 15.608 |
| 0.2866 | 16.67 | 5600 | 0.3426 | 43.3565 | 7365 | 15.7112 |
| 0.2817 | 17.86 | 6000 | 0.3434 | 43.6685 | 7365 | 15.6191 |
| 0.2802 | 19.05 | 6400 | 0.3423 | 43.6242 | 7365 | 15.6804 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.0+cu111
- Datasets 2.5.1
- Tokenizers 0.12.1
