metadata
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: t5-russian-spell
results: []
t5-russian-spell
This model is a fine-tuned version of sberbank-ai/ruT5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4066
- Rouge1: 44.2214
- Rouge2: 21.688
- Rougel: 44.2793
- Rougelsum: 44.0781
- Gen Len: 60.87
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: 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: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 0.2958 | 0.2 | 2500 | 0.4393 | 43.9635 | 21.3982 | 43.9784 | 43.8423 | 61.338 |
| 0.2427 | 0.4 | 5000 | 0.4460 | 44.609 | 22.1448 | 44.6314 | 44.4817 | 61.028 |
| 0.5326 | 0.6 | 7500 | 0.4100 | 44.7071 | 21.9365 | 44.7491 | 44.5944 | 60.844 |
| 0.5262 | 0.8 | 10000 | 0.4066 | 44.2214 | 21.688 | 44.2793 | 44.0781 | 60.87 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6