t5-arxiv
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.3852
- Rouge1: 18.0722
- Rouge2: 6.8453
- Rougel: 14.3659
- Rougelsum: 16.4137
- Gen Len: 19.0
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.5169 | 1.0 | 12500 | 2.3852 | 18.0722 | 6.8453 | 14.3659 | 16.4137 | 19.0 |
Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1