|
--- |
|
license: apache-2.0 |
|
base_model: t5-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
- wer |
|
model-index: |
|
- name: t-5-base-baseline |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# t-5-base-baseline |
|
|
|
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.1670 |
|
- Rouge1: 0.6791 |
|
- Rouge2: 0.4136 |
|
- Rougel: 0.6183 |
|
- Rougelsum: 0.6185 |
|
- Wer: 0.4846 |
|
- Bleurt: 0.3314 |
|
|
|
## 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: 6 |
|
- eval_batch_size: 6 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 2 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Wer | Bleurt | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:------:|:------:| |
|
| No log | 0.14 | 250 | 1.3197 | 0.6548 | 0.3804 | 0.5904 | 0.5905 | 0.5188 | 0.3009 | |
|
| 1.7026 | 0.27 | 500 | 1.2676 | 0.6613 | 0.3904 | 0.5985 | 0.5987 | 0.5074 | 0.3009 | |
|
| 1.7026 | 0.41 | 750 | 1.2385 | 0.6664 | 0.3961 | 0.6043 | 0.6044 | 0.5023 | 0.3009 | |
|
| 1.3446 | 0.55 | 1000 | 1.2234 | 0.6691 | 0.4009 | 0.6075 | 0.6076 | 0.4972 | 0.3009 | |
|
| 1.3446 | 0.68 | 1250 | 1.2089 | 0.671 | 0.4031 | 0.6099 | 0.6101 | 0.4944 | 0.3009 | |
|
| 1.309 | 0.82 | 1500 | 1.1983 | 0.673 | 0.4064 | 0.6121 | 0.6123 | 0.4914 | 0.3009 | |
|
| 1.309 | 0.96 | 1750 | 1.1900 | 0.6744 | 0.4075 | 0.6136 | 0.6137 | 0.4897 | 0.3009 | |
|
| 1.2783 | 1.09 | 2000 | 1.1840 | 0.6744 | 0.4082 | 0.614 | 0.6141 | 0.4889 | 0.2798 | |
|
| 1.2783 | 1.23 | 2250 | 1.1808 | 0.6759 | 0.4102 | 0.6154 | 0.6156 | 0.4875 | 0.2944 | |
|
| 1.2683 | 1.36 | 2500 | 1.1763 | 0.6771 | 0.4111 | 0.6163 | 0.6165 | 0.4863 | 0.3153 | |
|
| 1.2683 | 1.5 | 2750 | 1.1720 | 0.6772 | 0.4115 | 0.6169 | 0.617 | 0.4859 | 0.4109 | |
|
| 1.2469 | 1.64 | 3000 | 1.1701 | 0.6783 | 0.4127 | 0.6176 | 0.6178 | 0.485 | 0.3314 | |
|
| 1.2469 | 1.77 | 3250 | 1.1680 | 0.6786 | 0.4131 | 0.6178 | 0.6179 | 0.4849 | 0.3314 | |
|
| 1.2171 | 1.91 | 3500 | 1.1670 | 0.6791 | 0.4136 | 0.6183 | 0.6185 | 0.4846 | 0.3314 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.2 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
