metadata
license: apache-2.0
base_model: google/flan-t5-small
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: t5-summarization-one-shot-better-prompt
results: []
t5-summarization-one-shot-better-prompt
This model is a fine-tuned version of google/flan-t5-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.4802
- Rouge: {'rouge1': 37.3827, 'rouge2': 17.5806, 'rougeL': 20.1333, 'rougeLsum': 20.1333}
- Bert Score: 0.881
- Bleurt 20: -0.8056
- Gen Len: 13.305
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: 0.0001
- train_batch_size: 7
- eval_batch_size: 7
- seed: 42
- 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 | Rouge | Bert Score | Bleurt 20 | Gen Len |
|---|---|---|---|---|---|---|---|
| 2.8911 | 1.0 | 172 | 2.6479 | {'rouge1': 44.1146, 'rouge2': 17.635, 'rougeL': 19.7288, 'rougeLsum': 19.7288} | 0.8758 | -0.8151 | 14.79 |
| 2.7068 | 2.0 | 344 | 2.5580 | {'rouge1': 42.4931, 'rouge2': 18.5074, 'rougeL': 19.757, 'rougeLsum': 19.757} | 0.8774 | -0.8319 | 14.025 |
| 2.4884 | 3.0 | 516 | 2.5123 | {'rouge1': 43.5811, 'rouge2': 19.0798, 'rougeL': 20.1143, 'rougeLsum': 20.1143} | 0.8794 | -0.7737 | 14.455 |
| 2.3827 | 4.0 | 688 | 2.4875 | {'rouge1': 42.0721, 'rouge2': 18.4704, 'rougeL': 20.0131, 'rougeLsum': 20.0131} | 0.8801 | -0.7754 | 14.175 |
| 2.345 | 5.0 | 860 | 2.4596 | {'rouge1': 43.7021, 'rouge2': 20.0234, 'rougeL': 20.0962, 'rougeLsum': 20.0962} | 0.8809 | -0.7379 | 14.325 |
| 2.2438 | 6.0 | 1032 | 2.4466 | {'rouge1': 41.0624, 'rouge2': 18.8098, 'rougeL': 19.8672, 'rougeLsum': 19.8672} | 0.8803 | -0.7893 | 13.565 |
| 2.1878 | 7.0 | 1204 | 2.4505 | {'rouge1': 40.3802, 'rouge2': 18.9902, 'rougeL': 20.1633, 'rougeLsum': 20.1633} | 0.88 | -0.7735 | 13.26 |
| 2.084 | 8.0 | 1376 | 2.4384 | {'rouge1': 38.4615, 'rouge2': 18.3148, 'rougeL': 19.61, 'rougeLsum': 19.61} | 0.8802 | -0.7813 | 13.235 |
| 2.096 | 9.0 | 1548 | 2.4380 | {'rouge1': 38.9264, 'rouge2': 18.2137, 'rougeL': 19.7464, 'rougeLsum': 19.7464} | 0.8794 | -0.8013 | 13.18 |
| 2.0251 | 10.0 | 1720 | 2.4445 | {'rouge1': 36.7486, 'rouge2': 17.2998, 'rougeL': 20.0546, 'rougeLsum': 20.0546} | 0.8807 | -0.8057 | 12.97 |
| 2.0139 | 11.0 | 1892 | 2.4449 | {'rouge1': 38.374, 'rouge2': 18.4603, 'rougeL': 20.3794, 'rougeLsum': 20.3794} | 0.88 | -0.7885 | 13.415 |
| 1.957 | 12.0 | 2064 | 2.4502 | {'rouge1': 38.5085, 'rouge2': 18.9339, 'rougeL': 20.2576, 'rougeLsum': 20.2576} | 0.8812 | -0.7956 | 13.24 |
| 1.8508 | 13.0 | 2236 | 2.4510 | {'rouge1': 37.1308, 'rouge2': 17.4541, 'rougeL': 19.7904, 'rougeLsum': 19.7904} | 0.8802 | -0.7981 | 13.28 |
| 1.9315 | 14.0 | 2408 | 2.4590 | {'rouge1': 37.1896, 'rouge2': 17.8766, 'rougeL': 20.0, 'rougeLsum': 20.0} | 0.8813 | -0.7869 | 13.205 |
| 1.8447 | 15.0 | 2580 | 2.4635 | {'rouge1': 38.7071, 'rouge2': 18.5335, 'rougeL': 20.5302, 'rougeLsum': 20.5302} | 0.8813 | -0.7887 | 13.505 |
| 1.8488 | 16.0 | 2752 | 2.4643 | {'rouge1': 38.0242, 'rouge2': 17.8906, 'rougeL': 20.3013, 'rougeLsum': 20.3013} | 0.8818 | -0.7789 | 13.23 |
| 1.7909 | 17.0 | 2924 | 2.4739 | {'rouge1': 37.1561, 'rouge2': 17.7929, 'rougeL': 20.0763, 'rougeLsum': 20.0763} | 0.8806 | -0.7882 | 13.22 |
| 1.8615 | 18.0 | 3096 | 2.4770 | {'rouge1': 37.2891, 'rouge2': 17.6695, 'rougeL': 19.951, 'rougeLsum': 19.951} | 0.8803 | -0.8131 | 13.305 |
| 1.7938 | 19.0 | 3268 | 2.4796 | {'rouge1': 37.1339, 'rouge2': 17.6046, 'rougeL': 20.1063, 'rougeLsum': 20.1063} | 0.881 | -0.8031 | 13.26 |
| 1.7814 | 20.0 | 3440 | 2.4802 | {'rouge1': 37.3827, 'rouge2': 17.5806, 'rougeL': 20.1333, 'rougeLsum': 20.1333} | 0.881 | -0.8056 | 13.305 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2