Noisy-deberta-v3-xsmall-Label_B-768-epochs-5

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

• Loss: 0.0850
• Accuracy: 0.9839
• F1: 0.9839
• Precision: 0.9841
• Recall: 0.9839

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: 12
• eval_batch_size: 12
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 48
• 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_steps: 500
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.0801	0.9995	1066	0.1455	0.9542	0.9535	0.9565	0.9542
0.0581	1.9993	2132	0.0792	0.9805	0.9805	0.9807	0.9805
0.0543	2.9991	3198	0.3059	0.9434	0.9423	0.9495	0.9434
0.0003	3.9998	4265	0.0850	0.9839	0.9839	0.9841	0.9839
0.0006	4.9986	5330	0.1618	0.9737	0.9737	0.9747	0.9737

Framework versions

• Transformers 4.46.3
• Pytorch 2.4.0
• Datasets 3.1.0
• Tokenizers 0.20.3