|
--- |
|
license: apache-2.0 |
|
base_model: facebook/bart-base |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: Bart-base-v4 |
|
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-base-v4 |
|
|
|
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: 0.0390 |
|
|
|
## 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: 4 |
|
- total_train_batch_size: 16 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 6 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 0.095 | 0.32 | 250 | 0.0445 | |
|
| 0.0509 | 0.64 | 500 | 0.0425 | |
|
| 0.0495 | 0.96 | 750 | 0.0411 | |
|
| 0.0456 | 1.28 | 1000 | 0.0403 | |
|
| 0.0431 | 1.61 | 1250 | 0.0401 | |
|
| 0.0427 | 1.93 | 1500 | 0.0392 | |
|
| 0.0395 | 2.25 | 1750 | 0.0397 | |
|
| 0.0384 | 2.57 | 2000 | 0.0391 | |
|
| 0.0382 | 2.89 | 2250 | 0.0390 | |
|
| 0.0358 | 3.21 | 2500 | 0.0394 | |
|
| 0.0349 | 3.53 | 2750 | 0.0392 | |
|
| 0.0344 | 3.85 | 3000 | 0.0390 | |
|
| 0.0333 | 4.17 | 3250 | 0.0392 | |
|
| 0.0314 | 4.49 | 3500 | 0.0393 | |
|
| 0.0321 | 4.82 | 3750 | 0.0390 | |
|
| 0.0308 | 5.14 | 4000 | 0.0392 | |
|
| 0.0297 | 5.46 | 4250 | 0.0394 | |
|
| 0.0299 | 5.78 | 4500 | 0.0390 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.39.3 |
|
- Pytorch 2.1.2 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.15.2 |
|
|