Commit
•
cd48a25
1
Parent(s):
4b6e94e
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
model-index:
|
| 6 |
+
- name: bart-dnc-booksum
|
| 7 |
+
results: []
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 11 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 12 |
+
|
| 13 |
+
# bart-dnc-booksum
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [facebook/bart-large](https://huggingface.co/facebook/bart-large) on the None dataset.
|
| 16 |
+
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 6.5809
|
| 18 |
+
|
| 19 |
+
## Model description
|
| 20 |
+
|
| 21 |
+
More information needed
|
| 22 |
+
|
| 23 |
+
## Intended uses & limitations
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Training and evaluation data
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training procedure
|
| 32 |
+
|
| 33 |
+
### Training hyperparameters
|
| 34 |
+
|
| 35 |
+
The following hyperparameters were used during training:
|
| 36 |
+
- learning_rate: 5e-05
|
| 37 |
+
- train_batch_size: 1
|
| 38 |
+
- eval_batch_size: 1
|
| 39 |
+
- seed: 42
|
| 40 |
+
- gradient_accumulation_steps: 8
|
| 41 |
+
- total_train_batch_size: 8
|
| 42 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 43 |
+
- lr_scheduler_type: linear
|
| 44 |
+
- lr_scheduler_warmup_steps: 40
|
| 45 |
+
- num_epochs: 5
|
| 46 |
+
- mixed_precision_training: Native AMP
|
| 47 |
+
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 52 |
+
| 0.7554 | 0.16 | 150 | 4.3521 |
|
| 53 |
+
| 0.3115 | 0.32 | 300 | 4.9631 |
|
| 54 |
+
| 0.3536 | 0.48 | 450 | 4.8518 |
|
| 55 |
+
| 0.3419 | 0.64 | 600 | 4.8791 |
|
| 56 |
+
| 0.2082 | 0.8 | 750 | 5.8989 |
|
| 57 |
+
| 0.3123 | 0.97 | 900 | 4.8499 |
|
| 58 |
+
| 0.1517 | 1.13 | 1050 | 5.3516 |
|
| 59 |
+
| 0.2184 | 1.29 | 1200 | 5.5229 |
|
| 60 |
+
| 0.217 | 1.45 | 1350 | 5.7270 |
|
| 61 |
+
| 0.1375 | 1.61 | 1500 | 5.5408 |
|
| 62 |
+
| 0.1506 | 1.77 | 1650 | 5.3878 |
|
| 63 |
+
| 0.1904 | 1.93 | 1800 | 5.0405 |
|
| 64 |
+
| 0.0956 | 2.09 | 1950 | 6.0605 |
|
| 65 |
+
| 0.0856 | 2.25 | 2100 | 6.1505 |
|
| 66 |
+
| 0.1206 | 2.41 | 2250 | 5.4903 |
|
| 67 |
+
| 0.0872 | 2.58 | 2400 | 5.4324 |
|
| 68 |
+
| 0.1456 | 2.74 | 2550 | 6.1430 |
|
| 69 |
+
| 0.0848 | 2.9 | 2700 | 5.3999 |
|
| 70 |
+
| 0.079 | 3.06 | 2850 | 6.0992 |
|
| 71 |
+
| 0.0706 | 3.22 | 3000 | 6.2144 |
|
| 72 |
+
| 0.0768 | 3.38 | 3150 | 6.2338 |
|
| 73 |
+
| 0.0643 | 3.54 | 3300 | 6.0533 |
|
| 74 |
+
| 0.0576 | 3.7 | 3450 | 5.9912 |
|
| 75 |
+
| 0.0859 | 3.86 | 3600 | 6.3768 |
|
| 76 |
+
| 0.0615 | 4.02 | 3750 | 6.1106 |
|
| 77 |
+
| 0.072 | 4.19 | 3900 | 6.5398 |
|
| 78 |
+
| 0.0683 | 4.35 | 4050 | 6.5556 |
|
| 79 |
+
| 0.077 | 4.51 | 4200 | 6.6243 |
|
| 80 |
+
| 0.0811 | 4.67 | 4350 | 6.5221 |
|
| 81 |
+
| 0.0591 | 4.83 | 4500 | 6.5732 |
|
| 82 |
+
| 0.0699 | 4.99 | 4650 | 6.5809 |
|
| 83 |
+
|
| 84 |
+
|
| 85 |
+
### Framework versions
|
| 86 |
+
|
| 87 |
+
- Transformers 4.28.0
|
| 88 |
+
- Pytorch 2.0.1+cu118
|
| 89 |
+
- Datasets 2.14.4
|
| 90 |
+
- Tokenizers 0.13.3
|