File size: 3,172 Bytes
48dfe86 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 |
---
license: apache-2.0
base_model: facebook/bart-large
tags:
- generated_from_trainer
metrics:
- rouge
- wer
model-index:
- name: bart_bertsum_1024_375_1000
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bart_bertsum_1024_375_1000
This model is a fine-tuned version of [facebook/bart-large](https://huggingface.co./facebook/bart-large) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0535
- Rouge1: 0.6801
- Rouge2: 0.4119
- Rougel: 0.6159
- Rougelsum: 0.616
- Wer: 0.4729
- Bleurt: -0.3664
## 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: 2
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Wer | Bleurt |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:------:|:-------:|
| No log | 0.13 | 250 | 1.2919 | 0.636 | 0.3519 | 0.567 | 0.567 | 0.5296 | -0.0182 |
| 2.2326 | 0.27 | 500 | 1.2002 | 0.6503 | 0.3707 | 0.5816 | 0.5817 | 0.5113 | -0.7073 |
| 2.2326 | 0.4 | 750 | 1.1735 | 0.6564 | 0.3791 | 0.5898 | 0.5898 | 0.5048 | -0.3421 |
| 1.2886 | 0.53 | 1000 | 1.1476 | 0.661 | 0.3843 | 0.594 | 0.5939 | 0.4994 | 0.0835 |
| 1.2886 | 0.66 | 1250 | 1.1289 | 0.6615 | 0.3863 | 0.5938 | 0.5938 | 0.4945 | -0.5247 |
| 1.2306 | 0.8 | 1500 | 1.1197 | 0.67 | 0.3952 | 0.6046 | 0.6045 | 0.4909 | -0.192 |
| 1.2306 | 0.93 | 1750 | 1.1077 | 0.6734 | 0.3989 | 0.6068 | 0.6067 | 0.4876 | -0.3867 |
| 1.1852 | 1.06 | 2000 | 1.0917 | 0.6731 | 0.4027 | 0.609 | 0.609 | 0.4833 | -0.6453 |
| 1.1852 | 1.2 | 2250 | 1.0852 | 0.6707 | 0.4013 | 0.6054 | 0.6054 | 0.4824 | -0.5589 |
| 1.0875 | 1.33 | 2500 | 1.0785 | 0.6738 | 0.4049 | 0.6096 | 0.6096 | 0.4794 | -0.5107 |
| 1.0875 | 1.46 | 2750 | 1.0709 | 0.6743 | 0.4046 | 0.6096 | 0.6095 | 0.478 | -0.3387 |
| 1.0857 | 1.6 | 3000 | 1.0627 | 0.6778 | 0.41 | 0.6137 | 0.6137 | 0.4757 | -0.4275 |
| 1.0857 | 1.73 | 3250 | 1.0636 | 0.675 | 0.4088 | 0.6121 | 0.612 | 0.4745 | -0.3664 |
| 1.0634 | 1.86 | 3500 | 1.0552 | 0.6775 | 0.4103 | 0.6136 | 0.6136 | 0.4729 | -0.3664 |
| 1.0634 | 1.99 | 3750 | 1.0535 | 0.6801 | 0.4119 | 0.6159 | 0.616 | 0.4729 | -0.3664 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
|