metadata
base_model: kravchenko/uk-mt5-base
tags:
- summarization
- generated_from_trainer
datasets:
- xlsum
metrics:
- rouge
model-index:
- name: uk-mt5-base-xlsum-v3
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: xlsum
type: xlsum
config: ukrainian
split: validation
args: ukrainian
metrics:
- name: Rouge1
type: rouge
value: 4.0823
uk-mt5-base-xlsum-v3
This model is a fine-tuned version of kravchenko/uk-mt5-base on the xlsum dataset. It achieves the following results on the evaluation set:
- Loss: 1.7752
- Rouge1: 4.0823
- Rouge2: 0.8039
- Rougel: 4.0033
- Rougelsum: 4.0125
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: 5.6e-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: 9
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 2.8372 | 1.0 | 3413 | 1.9145 | 2.8041 | 0.6205 | 2.8045 | 2.8124 |
| 2.3863 | 2.0 | 6826 | 1.8516 | 3.3759 | 0.5435 | 3.3257 | 3.3414 |
| 2.2434 | 3.0 | 10239 | 1.8298 | 3.6535 | 0.6412 | 3.6527 | 3.6437 |
| 2.2593 | 4.0 | 13652 | 1.8232 | 3.7021 | 0.6361 | 3.6648 | 3.685 |
| 2.2084 | 5.0 | 17065 | 1.8081 | 3.8312 | 0.6814 | 3.7756 | 3.7865 |
| 2.1702 | 6.0 | 20478 | 1.8012 | 3.7021 | 0.6141 | 3.6663 | 3.6728 |
| 2.1449 | 7.0 | 23891 | 1.7901 | 3.88 | 0.6412 | 3.8275 | 3.8328 |
| 2.1032 | 8.0 | 27304 | 1.7788 | 3.8141 | 0.7171 | 3.7729 | 3.761 |
| 2.0779 | 9.0 | 30717 | 1.7752 | 4.0823 | 0.8039 | 4.0033 | 4.0125 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1