metadata
license: apache-2.0
base_model: t5-small
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: ft-t5-small-on-info-lg
results: []
ft-t5-small-on-info-lg
This model is a fine-tuned version of t5-small on the opus100 dataset. It achieves the following results on the evaluation set:
- Loss: 0.5870
- Bleu: 0.3242
- Gen Len: 15.9841
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
Training results
Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
---|---|---|---|---|---|
No log | 1.0 | 177 | 0.6138 | 0.2725 | 15.5828 |
No log | 2.0 | 354 | 0.6061 | 0.2603 | 16.3376 |
0.6269 | 3.0 | 531 | 0.6008 | 0.2719 | 15.2102 |
0.6269 | 4.0 | 708 | 0.5975 | 0.2875 | 16.6847 |
0.6269 | 5.0 | 885 | 0.5946 | 0.2719 | 15.4013 |
0.598 | 6.0 | 1062 | 0.5927 | 0.2497 | 15.9427 |
0.598 | 7.0 | 1239 | 0.5908 | 0.2555 | 16.2675 |
0.598 | 8.0 | 1416 | 0.5899 | 0.2953 | 16.9936 |
0.5825 | 9.0 | 1593 | 0.5889 | 0.3467 | 17.2134 |
0.5825 | 10.0 | 1770 | 0.5881 | 0.3013 | 16.1242 |
0.5825 | 11.0 | 1947 | 0.5873 | 0.3261 | 15.551 |
0.5695 | 12.0 | 2124 | 0.5871 | 0.2874 | 15.3854 |
0.5695 | 13.0 | 2301 | 0.5868 | 0.2987 | 15.5446 |
0.5695 | 14.0 | 2478 | 0.5869 | 0.3124 | 15.9013 |
0.5618 | 15.0 | 2655 | 0.5870 | 0.3242 | 15.9841 |
Framework versions
- Transformers 4.42.3
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1