Fill-Mask
Transformers
PyTorch
TensorBoard
Safetensors
Portuguese
xlm-roberta
Generated from Trainer
# Load model directly
from transformers import AutoTokenizer, AutoModelForMaskedLM
tokenizer = AutoTokenizer.from_pretrained("Luciano/xlm-roberta-base-finetuned-lener_br")
model = AutoModelForMaskedLM.from_pretrained("Luciano/xlm-roberta-base-finetuned-lener_br")Quick Links
xlm-roberta-base-finetuned-lener_br
This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9094
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.545 | 1.0 | 2079 | 1.4107 |
| 1.4708 | 2.0 | 4158 | 1.3126 |
| 1.322 | 3.0 | 6237 | 1.1943 |
| 1.1986 | 4.0 | 8316 | 1.1581 |
| 1.1316 | 5.0 | 10395 | 1.1156 |
| 1.0824 | 6.0 | 12474 | 1.0400 |
| 1.0435 | 7.0 | 14553 | 1.0276 |
| 0.9824 | 8.0 | 16632 | 1.0119 |
| 0.9289 | 9.0 | 18711 | nan |
| 0.9123 | 10.0 | 20790 | 0.9945 |
| 0.8591 | 11.0 | 22869 | nan |
| 0.8411 | 12.0 | 24948 | 0.9413 |
| 0.8376 | 13.0 | 27027 | 0.9411 |
| 0.7868 | 14.0 | 29106 | 0.9228 |
| 0.8012 | 15.0 | 31185 | 0.9449 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.1
- Downloads last month
- 3
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="Luciano/xlm-roberta-base-finetuned-lener_br")