defect-classification-distilbert-baseline-25-epochs

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

• Loss: 0.2683
• Accuracy: 0.8834

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: 2e-05
• train_batch_size: 512
• eval_batch_size: 512
• 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
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6829	1.0	1062	0.5554	0.7841
0.4905	2.0	2124	0.4389	0.8172
0.4418	3.0	3186	0.3930	0.8379
0.4167	4.0	4248	0.3503	0.8491
0.4166	5.0	5310	0.3163	0.8612
0.4344	6.0	6372	0.3135	0.8638
0.3547	7.0	7434	0.3092	0.8648
0.4277	8.0	8496	0.3099	0.8633
0.399	9.0	9558	0.3071	0.8660
0.4125	10.0	10620	0.2843	0.8781
0.3662	11.0	11682	0.2899	0.8736
0.3396	12.0	12744	0.2796	0.8782
0.3775	13.0	13806	0.2797	0.8803
0.3552	14.0	14868	0.2757	0.8815
0.3208	15.0	15930	0.2747	0.8807
0.3344	16.0	16992	0.2702	0.8839
0.3171	17.0	18054	0.2745	0.8782
0.3535	18.0	19116	0.2745	0.8799
0.394	19.0	20178	0.2669	0.8866
0.299	20.0	21240	0.2720	0.8804
0.3209	21.0	22302	0.2720	0.8790
0.3366	22.0	23364	0.2696	0.8818
0.3531	23.0	24426	0.2690	0.8826
0.3368	24.0	25488	0.2685	0.8826
0.3251	25.0	26550	0.2683	0.8834

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0