metadata
license: mit
tags:
- generated_from_trainer
- nlu
- intent-classification
- text-classification
metrics:
- accuracy
- f1
model-index:
- name: xlm-r-base-amazon-massive-intent-label_smoothing
results:
- task:
name: intent-classification
type: intent-classification
dataset:
name: MASSIVE
type: AmazonScience/massive
split: test
metrics:
- name: F1
type: f1
value: 0.8879
datasets:
- AmazonScience/massive
language:
- en
xlm-r-base-amazon-massive-intent-label_smoothing
This model is a fine-tuned version of xlm-roberta-base on the MASSIVE1.1 dataset. It achieves the following results on the evaluation set:
- Loss: 2.5148
- Accuracy: 0.8879
- F1: 0.8879
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- label_smoothing_factor: 0.4
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 3.3945 | 1.0 | 720 | 2.7175 | 0.7900 | 0.7900 |
| 2.7629 | 2.0 | 1440 | 2.5660 | 0.8549 | 0.8549 |
| 2.5143 | 3.0 | 2160 | 2.5389 | 0.8711 | 0.8711 |
| 2.4678 | 4.0 | 2880 | 2.5172 | 0.8883 | 0.8883 |
| 2.4187 | 5.0 | 3600 | 2.5148 | 0.8879 | 0.8879 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.0
- Tokenizers 0.13.2