metaphor-cat-roberta-large-weights

This model is a fine-tuned version of projecte-aina/roberta-large-ca-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2147
• Precision: 0.7879
• Recall: 0.6190
• F1: 0.6933
• Accuracy: 0.9725

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: 3e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	66	0.6639	0.3276	0.4524	0.3800	0.9258
0.619	2.0	132	0.3678	0.3976	0.7857	0.528	0.9294
0.619	3.0	198	0.3450	0.4487	0.8333	0.5833	0.9402
0.1982	4.0	264	1.0043	0.8710	0.6429	0.7397	0.9773
0.0613	5.0	330	0.9192	0.6829	0.6667	0.6747	0.9677
0.0613	6.0	396	1.0697	0.8125	0.6190	0.7027	0.9737
0.0331	7.0	462	1.2183	0.8125	0.6190	0.7027	0.9737
0.0238	8.0	528	1.1401	0.8182	0.6429	0.7200	0.9749
0.0238	9.0	594	1.2380	0.8065	0.5952	0.6849	0.9725
0.0191	10.0	660	1.2187	0.8125	0.6190	0.7027	0.9737
0.0098	11.0	726	1.3181	0.8065	0.5952	0.6849	0.9725
0.0098	12.0	792	1.3824	0.8065	0.5952	0.6849	0.9725
0.0063	13.0	858	1.3370	0.8333	0.5952	0.6944	0.9737
0.0048	14.0	924	1.2230	0.7879	0.6190	0.6933	0.9725
0.0048	15.0	990	1.2147	0.7879	0.6190	0.6933	0.9725

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1