polar4

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5187
• Accuracy: 0.7426
• F1: 0.7199
• Precision: 0.7477
• Recall: 0.7426

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: 0.0002
• train_batch_size: 200
• eval_batch_size: 200
• seed: 42
• 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_ratio: 0.2
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.6359	7.6923	100	0.6229	0.6357	0.4941	0.4041	0.6357
0.5953	15.3846	200	0.5654	0.7039	0.6698	0.7030	0.7039
0.5605	23.0769	300	0.5510	0.6930	0.6297	0.7259	0.6930
0.5578	30.7692	400	0.5273	0.7426	0.7384	0.7376	0.7426
0.5635	38.4615	500	0.5198	0.7364	0.7257	0.7301	0.7364
0.5443	46.1538	600	0.5214	0.7271	0.6978	0.7338	0.7271

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.4.1
• Tokenizers 0.22.1