computer_parts_classifier-model

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.5117
• Accuracy: 0.8138

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.97	9	1.0525	0.5517
1.0645	1.95	18	0.9405	0.6
0.9405	2.92	27	0.7902	0.7034
0.7669	4.0	37	0.6923	0.7379
0.6008	4.97	46	0.6152	0.7862
0.5142	5.95	55	0.5639	0.7931
0.394	6.92	64	0.5640	0.8
0.3649	8.0	74	0.5181	0.7862
0.279	8.97	83	0.5094	0.8345
0.2549	9.95	92	0.4882	0.8276
0.1925	10.92	101	0.5041	0.8
0.2185	12.0	111	0.5195	0.8138
0.1921	12.97	120	0.5170	0.8
0.1921	13.95	129	0.5846	0.7793
0.15	14.92	138	0.5217	0.8207
0.1798	16.0	148	0.5421	0.7862
0.1729	16.97	157	0.5516	0.8207
0.1459	17.95	166	0.5438	0.7931
0.1701	18.92	175	0.5043	0.8345
0.1487	19.46	180	0.5117	0.8138

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2