roberta-decision-model-11classes-v2

This model is a fine-tuned version of roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0019
• Accuracy: 0.9998
• F1: 0.9997

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: 32
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• 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: 0.06
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.3921	0.1818	500	0.0324	0.9978	0.9978
0.0363	0.3636	1000	0.0074	0.9995	0.9995
0.0285	0.5455	1500	0.0047	0.9995	0.9995
0.0038	0.7273	2000	0.0023	0.9997	0.9997
0.0192	0.9091	2500	0.0021	0.9998	0.9997
0.0061	1.0909	3000	0.0032	0.9997	0.9997
0.0181	1.2727	3500	0.0024	0.9997	0.9997
0.0151	1.4545	4000	0.0033	0.9997	0.9997
0.0042	1.6364	4500	0.0025	0.9998	0.9997
0.0031	1.8182	5000	0.0020	0.9998	0.9997
0.0028	2.0	5500	0.0020	0.9998	0.9997
0.0028	2.1818	6000	0.0021	0.9998	0.9997
0.0042	2.3636	6500	0.0023	0.9998	0.9997
0.0099	2.5455	7000	0.0020	0.9998	0.9997
0.0014	2.7273	7500	0.0020	0.9998	0.9997
0.0013	2.9091	8000	0.0019	0.9998	0.9997

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2