metadata
license: mit
base_model: sn4kebyt3/ru-bart-large
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: bart-rebuilder
results: []
bart-rebuilder
This model is a fine-tuned version of sn4kebyt3/ru-bart-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8215
- Bleu: 50.6678
- Chrf: 77.7388
- Gen Len: 23.5257
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: 0.0003
- 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: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf | Gen Len |
|---|---|---|---|---|---|---|
| 0.5679 | 1.0 | 6914 | 0.7178 | 45.1359 | 74.9569 | 23.4132 |
| 0.2277 | 2.0 | 13828 | 0.6731 | 49.1274 | 76.9144 | 23.4925 |
| 0.0424 | 3.0 | 20742 | 0.8215 | 50.6678 | 77.7388 | 23.5257 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2