roberta-base-continual-kennedy2020constructing-S0T1

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

• Loss: 1.5398
• Accuracy: 0.5667
• Roc Auc: 0.7793
• Micro Precision: 0.5667
• Macro Precision: 0.5834
• Weighted Precision: 0.5834
• Micro Recall: 0.5667
• Macro Recall: 0.5667
• Weighted Recall: 0.5667
• Micro F1: 0.5667
• Macro F1: 0.5568
• Weighted F1: 0.5568

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Roc Auc	Micro Precision	Macro Precision	Weighted Precision	Micro Recall	Macro Recall	Weighted Recall	Micro F1	Macro F1	Weighted F1
No log	1.0	57	1.3397	0.3333	0.6652	0.3333	0.1111	0.1111	0.3333	0.3333	0.3333	0.3333	0.1667	0.1667
No log	2.0	114	1.0461	0.48	0.7563	0.48	0.3508	0.3508	0.48	0.48	0.48	0.48	0.3810	0.3810
No log	3.0	171	1.0020	0.5533	0.7707	0.5533	0.4876	0.4876	0.5533	0.5533	0.5533	0.5533	0.4667	0.4667
No log	4.0	228	1.1591	0.5133	0.7707	0.5133	0.5582	0.5582	0.5133	0.5133	0.5133	0.5133	0.4790	0.4790
No log	5.0	285	1.2874	0.5467	0.7651	0.5467	0.5471	0.5471	0.5467	0.5467	0.5467	0.5467	0.5278	0.5278
No log	6.0	342	1.1093	0.6067	0.7936	0.6067	0.5863	0.5863	0.6067	0.6067	0.6067	0.6067	0.5874	0.5874
No log	7.0	399	1.2420	0.6	0.7948	0.6	0.5925	0.5925	0.6	0.6	0.6	0.6	0.5860	0.5860
No log	8.0	456	1.5520	0.5467	0.7737	0.5467	0.5457	0.5457	0.5467	0.5467	0.5467	0.5467	0.5282	0.5282
0.4501	9.0	513	1.5398	0.5667	0.7793	0.5667	0.5834	0.5834	0.5667	0.5667	0.5667	0.5667	0.5568	0.5568

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0