metadata
base_model: ai-forever/ruT5-base
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: skilltext
results: []
skilltext
This model is a fine-tuned version of ai-forever/ruT5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.0577
- Rouge1: 30.9205
- Rouge2: 11.9258
- Rougel: 26.6497
- Rougelsum: 26.4407
- Gen Len: 18.6875
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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| No log | 1.6129 | 50 | 1.8405 | 16.798 | 2.7473 | 15.7599 | 15.6863 | 19.0 |
| No log | 3.2258 | 100 | 1.4606 | 19.3942 | 7.4911 | 18.9911 | 18.8407 | 18.875 |
| No log | 4.8387 | 150 | 1.3583 | 27.0146 | 13.8805 | 24.3188 | 24.2122 | 18.6875 |
| No log | 6.4516 | 200 | 1.2490 | 32.855 | 15.9819 | 31.0776 | 30.8624 | 18.75 |
| No log | 8.0645 | 250 | 1.1590 | 30.3762 | 11.8253 | 27.5559 | 27.2332 | 18.5625 |
| No log | 9.6774 | 300 | 1.1469 | 37.2275 | 17.107 | 33.4177 | 33.3688 | 18.4375 |
| No log | 11.2903 | 350 | 1.1364 | 34.3596 | 15.6845 | 30.8838 | 31.0842 | 18.625 |
| No log | 12.9032 | 400 | 1.0927 | 34.9322 | 15.8027 | 30.2917 | 30.1379 | 18.6875 |
| No log | 14.5161 | 450 | 1.0672 | 32.2753 | 15.7727 | 28.1883 | 27.8978 | 18.6875 |
| 1.8948 | 16.1290 | 500 | 1.0721 | 37.6573 | 15.6507 | 32.7817 | 32.742 | 18.5625 |
| 1.8948 | 17.7419 | 550 | 1.0692 | 34.958 | 15.3422 | 30.4656 | 30.3306 | 18.5 |
| 1.8948 | 19.3548 | 600 | 1.0577 | 30.9205 | 11.9258 | 26.6497 | 26.4407 | 18.6875 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.2
- Datasets 2.12.0
- Tokenizers 0.19.1