metadata
license: apache-2.0
base_model: facebook/bart-large
tags:
- generated_from_trainer
model-index:
- name: bart_baseline_peft
results: []
bart_baseline_peft
This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.0163
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: 0.0002
- 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: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0.18 | 250 | 1.4711 |
| 1.8588 | 0.35 | 500 | 1.4005 |
| 1.8588 | 0.53 | 750 | 1.3036 |
| 1.4092 | 0.71 | 1000 | 1.2629 |
| 1.4092 | 0.89 | 1250 | 1.2011 |
| 1.2502 | 1.06 | 1500 | 1.1580 |
| 1.2502 | 1.24 | 1750 | 1.1387 |
| 0.9815 | 1.42 | 2000 | 1.1045 |
| 0.9815 | 1.6 | 2250 | 1.0710 |
| 0.9213 | 1.77 | 2500 | 1.0347 |
| 0.9213 | 1.95 | 2750 | 1.0163 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2