computer_partsclassifier-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: 0.5569
• 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	1.0514	0.5172
1.0783	1.95	18	0.9347	0.6828
0.9676	2.92	27	0.7734	0.7517
0.7674	4.0	37	0.6470	0.7931
0.6162	4.97	46	0.5806	0.8
0.4838	5.95	55	0.5836	0.7931
0.4034	6.92	64	0.5778	0.8
0.325	8.0	74	0.5584	0.8069
0.2824	8.97	83	0.4549	0.8207
0.2252	9.95	92	0.5479	0.8
0.2017	10.92	101	0.5885	0.7724
0.183	12.0	111	0.5698	0.8
0.1709	12.97	120	0.5687	0.8
0.1709	13.95	129	0.6270	0.7793
0.1647	14.92	138	0.5652	0.8
0.1543	16.0	148	0.5965	0.8138
0.1676	16.97	157	0.5710	0.8
0.1562	17.95	166	0.6193	0.7724
0.1402	18.92	175	0.6086	0.7862
0.1313	19.46	180	0.5569	0.8138

Framework versions

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