Downloads · 30 days
10
6% of all-time downloads
ukrt5-formality-transfer-ukrainian-v4
ViHr/ukrt5-formality-transfer-ukrainian-v4
ukrt5-formality-transfer-ukrainian-v4 is a machine learning model from ViHr. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as mit.
should probably proofread and complete it, then remove this comment. --
All-time downloads
169
Public
Parameters
252M
2.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1 GB · 100%
At a glance
- Library
- transformers
- License
- mit
- Model type
- t5
- Access
- Public
- Created
- May 28, 2026
- Updated
- May 29, 2026
- SHA
- beb82067
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
ukrt5-formality-transfer-ukrainian-v4
This model is a fine-tuned version of uaritm/ukrt5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3509
- Bleu: 29.8446
- Mean Pred Words: 8.8865
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 0.05
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Mean Pred Words |
|---|---|---|---|---|---|
| 6.9646 | 0.2611 | 2000 | 2.3834 | 12.7318 | 11.2575 |
| 5.0468 | 0.5221 | 4000 | 1.8917 | 24.5780 | 9.2275 |
| 4.5162 | 0.7832 | 6000 | 1.7397 | 26.1762 | 9.0015 |
| 4.1834 | 1.0443 | 8000 | 1.6485 | 27.3852 | 9.011 |
| 3.9961 | 1.3053 | 10000 | 1.5766 | 27.8863 | 8.981 |
| 3.8509 | 1.5664 | 12000 | 1.5287 | 28.2907 | 8.9795 |
| 3.7421 | 1.8274 | 14000 | 1.4980 | 28.6894 | 8.926 |
| 3.6209 | 2.0885 | 16000 | 1.4653 | 28.8614 | 8.9455 |
| 3.5443 | 2.3496 | 18000 | 1.4389 | 28.9359 | 8.93 |
| 3.5049 | 2.6106 | 20000 | 1.4210 | 29.2368 | 8.8765 |
| 3.4444 | 2.8717 | 22000 | 1.4088 | 29.2687 | 8.9095 |
| 3.3854 | 3.1328 | 24000 | 1.3931 | 29.5429 | 8.897 |
| 3.3632 | 3.3938 | 26000 | 1.3801 | 29.4986 | 8.9135 |
| 3.3391 | 3.6549 | 28000 | 1.3725 | 29.6629 | 8.881 |
| 3.2867 | 3.9159 | 30000 | 1.3643 | 29.7246 | 8.9105 |
| 3.2754 | 4.1770 | 32000 | 1.3577 | 29.8422 | 8.897 |
| 3.2765 | 4.4381 | 34000 | 1.3546 | 29.8180 | 8.9015 |
| 3.2598 | 4.6991 | 36000 | 1.3519 | 29.8762 | 8.8915 |
| 3.2576 | 4.9602 | 38000 | 1.3509 | 29.8446 | 8.8865 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2
- Library
- transformers
- Type
- t5
- License
- mit
- Created
- May 28, 2026
- Updated
- May 29, 2026