metadata
license: apache-2.0
base_model: facebook/bart-large
tags:
- generated_from_trainer
metrics:
- rouge
- wer
model-index:
- name: bart_extractive_1024_750
results: []
bart_extractive_1024_750
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: 0.9368
- Rouge1: 0.7111
- Rouge2: 0.4588
- Rougel: 0.6541
- Rougelsum: 0.6542
- Wer: 0.433
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: 6
- eval_batch_size: 6
- 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 | Wer |
|---|---|---|---|---|---|---|---|---|
| No log | 0.13 | 250 | 1.1442 | 0.6749 | 0.4062 | 0.6133 | 0.6132 | 0.4806 |
| 2.053 | 0.27 | 500 | 1.0353 | 0.6859 | 0.4274 | 0.6269 | 0.6269 | 0.4586 |
| 2.053 | 0.4 | 750 | 1.0013 | 0.6935 | 0.4384 | 0.6351 | 0.6352 | 0.4499 |
| 1.1091 | 0.53 | 1000 | 0.9866 | 0.7003 | 0.4467 | 0.6416 | 0.6417 | 0.4425 |
| 1.1091 | 0.66 | 1250 | 0.9591 | 0.7052 | 0.4512 | 0.6469 | 0.647 | 0.4386 |
| 1.0491 | 0.8 | 1500 | 0.9502 | 0.7035 | 0.4517 | 0.6469 | 0.647 | 0.4366 |
| 1.0491 | 0.93 | 1750 | 0.9368 | 0.7111 | 0.4588 | 0.6541 | 0.6542 | 0.433 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2