File size: 1,968 Bytes
111d8e0 |
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 |
---
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: wiki_asp-software_5406_bart-base
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. -->
# wiki_asp-software_5406_bart-base
This model is a fine-tuned version of [facebook/bart-base](https://huggingface.co/facebook/bart-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.0105
- Rouge1: 0.1448
- Rouge2: 0.0489
- Rougel: 0.1205
- Rougelsum: 0.1207
- Gen Len: 19.8626
## 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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:-------:|
| No log | 2.37 | 500 | 3.1540 | 0.1377 | 0.0441 | 0.1151 | 0.1153 | 19.8858 |
| No log | 4.74 | 1000 | 3.0456 | 0.1408 | 0.0463 | 0.117 | 0.1172 | 19.7587 |
| No log | 7.1 | 1500 | 3.0225 | 0.1428 | 0.0472 | 0.1183 | 0.1183 | 19.9145 |
| 3.2197 | 9.47 | 2000 | 3.0105 | 0.1448 | 0.0489 | 0.1205 | 0.1207 | 19.8626 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.0.0+cu117
- Datasets 2.18.0
- Tokenizers 0.15.2
|