metadata

base_model: cardiffnlp/twitter-roberta-base-irony
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: twitter-roberta-base_3epoch10.16
    results: []

twitter-roberta-base_3epoch10.16

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

Loss: 2.1539
Accuracy: 0.7478
F1: 0.4224
Precision: 0.6154
Recall: 0.3216
Precision Sarcastic: 0.6154
Recall Sarcastic: 0.3216
F1 Sarcastic: 0.4224

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: 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Precision Sarcastic	Recall Sarcastic	F1 Sarcastic
No log	1.0	174	2.1148	0.7493	0.2869	0.7778	0.1759	0.7778	0.1759	0.2869
No log	2.0	348	1.2416	0.7320	0.3882	0.5619	0.2965	0.5619	0.2965	0.3882
0.0732	3.0	522	1.5725	0.7392	0.4873	0.5584	0.4322	0.5584	0.4322	0.4873
0.0732	4.0	696	1.7604	0.7450	0.4520	0.5887	0.3668	0.5887	0.3668	0.4520
0.0732	5.0	870	1.9529	0.7291	0.4749	0.5346	0.4271	0.5346	0.4271	0.4749
0.0278	6.0	1044	1.7258	0.7334	0.4699	0.5467	0.4121	0.5467	0.4121	0.4699
0.0278	7.0	1218	2.0437	0.7464	0.4568	0.592	0.3719	0.592	0.3719	0.4568
0.0278	8.0	1392	2.0771	0.7507	0.4508	0.6121	0.3568	0.6121	0.3568	0.4508
0.0081	9.0	1566	2.1414	0.7478	0.4186	0.6176	0.3166	0.6176	0.3166	0.4186
0.0081	10.0	1740	2.1539	0.7478	0.4224	0.6154	0.3216	0.6154	0.3216	0.4224

Framework versions

Transformers 4.41.2
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1