NiloofarMomeni/distilhubert-finetuned-roughness

NiloofarMomeni/distilhubert-finetuned-strain-finetuned-roughness

This model is a fine-tuned version of ntu-spml/distilhubert on the PQVD dataset. It achieves the following results on the evaluation set:

• Loss: 2.2467
• Accuracy: 0.7253

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: 16
• eval_batch_size: 16
• 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: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5013	1.0	46	1.0131	0.6593
0.4913	2.0	92	0.6931	0.6813
0.3487	3.0	138	0.8073	0.6044
0.3173	4.0	184	0.7272	0.6813
0.2558	5.0	230	1.0216	0.6593
0.2806	6.0	276	0.9014	0.6923
0.1703	7.0	322	1.1948	0.6484
0.1371	8.0	368	1.4679	0.6484
0.1612	9.0	414	1.4535	0.6484
0.1375	10.0	460	1.6594	0.6593
0.0171	11.0	506	1.9542	0.6484
0.0014	12.0	552	1.8910	0.7033
0.0017	13.0	598	2.0565	0.7033
0.0006	14.0	644	2.0231	0.7033
0.0006	15.0	690	2.1247	0.6923
0.0005	16.0	736	2.2004	0.6813
0.0003	17.0	782	2.2082	0.7033
0.0002	18.0	828	2.2396	0.7033
0.0002	19.0	874	2.2416	0.7253
0.0002	20.0	920	2.2467	0.7253

Framework versions

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