t5-base-msmarco-nlgen-cb
This model is a fine-tuned version of t5-base on the MS MARCO Natural Language Generation dataset. It achieves the following results on the evaluation set:
- Loss: 2.0571
- Rougelsum: 24.7427
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: 16
- 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
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rougelsum |
|---|---|---|---|---|
| 2.1393 | 0.26 | 2500 | 2.1099 | 24.5028 |
| 2.1006 | 0.52 | 5000 | 2.0739 | 24.6017 |
| 2.0694 | 0.78 | 7500 | 2.0571 | 24.7427 |
Framework versions
- Transformers 4.23.0.dev0
- Pytorch 1.12.1+cu102
- Datasets 2.4.0
- Tokenizers 0.12.1