|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: result |
|
results: [] |
|
--- |
|
|
|
This model is a fine-tuned version of [neuralmind/bert-large-portuguese-cased](https://huggingface.co/neuralmind/bert-large-portuguese-cased) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8031 |
|
|
|
## 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: 2 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3.0 |
|
|
|
### Training results |
|
|
|
### Eval metrics |
|
|
|
epoch = 3.0 |
|
eval_loss = 0.8031 |
|
eval_runtime = 0:02:17.77 |
|
eval_samples = 134 |
|
eval_samples_per_second = 0.973 |
|
eval_steps_per_second = 0.123 |
|
perplexity = 2.2325 |
|
|
|
### Framework versions |
|
|
|
- Transformers 4.13.0.dev0 |
|
- Pytorch 1.10.0+cu102 |
|
- Datasets 1.16.1 |
|
- Tokenizers 0.10.3 |
