README.md

metadata

license: mit
tags:
  - generated_from_trainer
metrics:
  - f1
  - precision
  - recall
model-index:
  - name: roberta-bne-clara
    results: []

roberta-bne-clara

This model is a fine-tuned version of roberta-bne on the Diario de Madrid dataset. It achieves the following results on the evaluation set:

• Loss: 0.2078
• F1: 0.5174
• Precision: 0.4561
• Recall: 0.5977

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: 12
• eval_batch_size: 12
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Precision	Recall
0.7026	1.0	103	0.2844	0.4720	0.5135	0.4368
0.325	2.0	206	0.1715	0.5665	0.5698	0.5632
0.2094	3.0	309	0.1452	0.6294	0.5636	0.7126
0.1416	4.0	412	0.1508	0.5178	0.4636	0.5862
0.1058	5.0	515	0.1794	0.5700	0.4917	0.6782
0.0711	6.0	618	0.1743	0.5510	0.4954	0.6207
0.0489	7.0	721	0.1900	0.5895	0.5437	0.6437
0.0373	8.0	824	0.1841	0.6131	0.5446	0.7011
0.0277	9.0	927	0.1954	0.5445	0.5	0.5977
0.0215	10.0	1030	0.2078	0.5174	0.4561	0.5977

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 1.16.1
• Tokenizers 0.13.2