t5-small-finetuned-BBCNews_v2
This model is a fine-tuned version of google-t5/t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3170
- Rouge1: 0.1558
- Rouge2: 0.1263
- Rougel: 0.1483
- Rougelsum: 0.1496
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: 4e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 5
- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 75 | 0.4430 | 0.1374 | 0.098 | 0.1257 | 0.1289 |
| No log | 1.99 | 150 | 0.3657 | 0.1466 | 0.1112 | 0.1367 | 0.1388 |
| No log | 2.99 | 225 | 0.3449 | 0.1536 | 0.1222 | 0.145 | 0.147 |
| No log | 3.99 | 300 | 0.3320 | 0.1534 | 0.1226 | 0.1454 | 0.147 |
| 0.609 | 5.0 | 376 | 0.3245 | 0.1534 | 0.1229 | 0.1457 | 0.1472 |
| 0.609 | 6.0 | 451 | 0.3214 | 0.155 | 0.125 | 0.147 | 0.1486 |
| 0.609 | 6.99 | 526 | 0.3181 | 0.1555 | 0.1261 | 0.148 | 0.1496 |
| 0.609 | 7.98 | 600 | 0.3170 | 0.1558 | 0.1263 | 0.1483 | 0.1496 |
Framework versions
- Transformers 4.32.0
- Pytorch 2.1.2
- Datasets 2.12.0
- Tokenizers 0.13.3