uas-nlp
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4366
- Rouge1: 42.1092
- Rouge2: 22.4228
- Rougel: 36.555
- Bleu: 20.5533
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 0.1
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Bleu |
|---|---|---|---|---|---|---|---|
| 2.2572 | 1.0 | 793 | 1.7467 | 28.3356 | 13.6831 | 25.3984 | 9.1195 |
| 1.8100 | 2.0 | 1586 | 1.6231 | 35.1004 | 17.8709 | 30.9105 | 15.0132 |
| 1.6857 | 3.0 | 2379 | 1.5609 | 36.8045 | 18.7151 | 32.167 | 16.0463 |
| 1.6135 | 4.0 | 3172 | 1.5192 | 38.0403 | 19.2669 | 33.1081 | 16.9419 |
| 1.5584 | 5.0 | 3965 | 1.5060 | 39.9211 | 20.6822 | 34.7901 | 18.7794 |
| 1.5097 | 6.0 | 4758 | 1.4814 | 40.6802 | 21.2563 | 35.2195 | 19.3699 |
| 1.4754 | 7.0 | 5551 | 1.4693 | 40.3298 | 21.5759 | 34.9535 | 19.7572 |
| 1.4389 | 8.0 | 6344 | 1.4640 | 40.9452 | 21.6119 | 35.1942 | 19.7244 |
| 1.4100 | 9.0 | 7137 | 1.4571 | 40.6946 | 21.3943 | 35.0856 | 19.8844 |
| 1.3894 | 10.0 | 7930 | 1.4501 | 40.8169 | 21.1663 | 35.5354 | 19.7837 |
| 1.3680 | 11.0 | 8723 | 1.4402 | 41.3369 | 21.8096 | 35.792 | 20.2212 |
| 1.3493 | 12.0 | 9516 | 1.4365 | 41.5085 | 21.9571 | 36.2856 | 20.3886 |
| 1.3328 | 13.0 | 10309 | 1.4447 | 41.051 | 21.3352 | 35.6354 | 19.6375 |
| 1.3184 | 14.0 | 11102 | 1.4445 | 41.9338 | 21.7355 | 35.5373 | 20.2003 |
| 1.3084 | 15.0 | 11895 | 1.4366 | 42.1092 | 22.4228 | 36.555 | 20.5533 |
Framework versions
- Transformers 5.10.2
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2