Edit model card

Clasificacion_sentimientos

This model is a fine-tuned version of BSC-TeMU/roberta-base-bne on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3399
  • Accuracy: 0.9428

Model description

Se entrena un modelo que es capaz de clasificar si es un comentario postivo o negativo.

Intended uses & limitations

More information needed

Training and evaluation data

Se entrenó el modelo usando comentarios de peliculas de la página $https://www.filmaffinity.com/es/main.html$

Training procedure

La variable review_rate se usó para clasificar los comentarios positivos y negativos así: Positivos: los rating con 8,9,10. Negativos: Los rating con 3,2,1.

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: 4

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.2566 1.0 901 0.5299 0.8935
0.0963 2.0 1802 0.2885 0.9383
0.0133 3.0 2703 0.3546 0.9406
0.0002 4.0 3604 0.3399 0.9428

Framework versions

  • Transformers 4.17.0
  • Pytorch 1.10.0+cu111
  • Datasets 2.0.0
  • Tokenizers 0.11.6
Downloads last month
14
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.