distilhubert-finetuned-PQVD

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: 0.0056
• Accuracy: 1.0

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• 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
0.4997	1.0	46	0.5211	0.6848
0.2498	2.0	92	0.1029	0.9891
0.3537	3.0	138	0.0987	0.9783
0.3084	4.0	184	0.1350	0.9457
0.0315	5.0	230	0.1213	0.9674
0.0036	6.0	276	0.0056	1.0
0.0886	7.0	322	0.0084	1.0
0.0014	8.0	368	0.0319	0.9891
0.0011	9.0	414	0.0018	1.0
0.0008	10.0	460	0.0122	0.9891
0.0005	11.0	506	0.0045	1.0
0.0004	12.0	552	0.0056	1.0
0.0004	13.0	598	0.0046	1.0
0.0004	14.0	644	0.0060	1.0
0.0004	15.0	690	0.0047	1.0
0.0003	16.0	736	0.0043	1.0
0.0004	17.0	782	0.0039	1.0
0.0003	18.0	828	0.0036	1.0
0.0003	19.0	874	0.0038	1.0
0.0002	20.0	920	0.0037	1.0

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2