|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: bert-base-uncased-Masked_Language_Model-US_Economic_News_Articles |
|
results: [] |
|
language: |
|
- en |
|
metrics: |
|
- perplexity |
|
--- |
|
|
|
# bert-base-uncased-Masked_Language_Model-US_Economic_News_Articles |
|
|
|
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased). |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.8322 |
|
|
|
## Model description |
|
|
|
This is a masked language modeling project. |
|
|
|
For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Masked%20Language%20Model/US%20Economic%20News%20Articles/US_Economic_News_Articles_MLM.ipynb |
|
|
|
## Intended uses & limitations |
|
|
|
This model is intended to demonstrate my ability to solve a complex problem using technology. |
|
|
|
## Training and evaluation data |
|
|
|
Dataset Source: https://www.kaggle.com/datasets/trikialaaa/2k-clean-medical-articles-medicalnewstoday |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 2e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 2.1833 | 1.0 | 2016 | 1.9529 | |
|
| 2.004 | 2.0 | 4032 | 1.9002 | |
|
| 1.941 | 3.0 | 6048 | 1.8600 | |
|
|
|
Perplexity: 6.25 |
|
|
|
### Framework versions |
|
|
|
- Transformers 4.27.1 |
|
- Pytorch 1.13.1+cu116 |
|
- Datasets 2.10.1 |
|
- Tokenizers 0.13.2 |
