car_orientation_classification_zoomed

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6108
• Accuracy: 0.7597

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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.9887	1.0	68	1.9011	0.3463
1.4388	2.0	136	1.3001	0.4594
1.1799	3.0	204	1.1267	0.4841
1.0245	4.0	272	0.9695	0.5936
0.8203	5.0	340	0.8157	0.6890
0.7146	6.0	408	0.7898	0.6678
0.6137	7.0	476	0.6343	0.7420
0.5746	8.0	544	0.6351	0.7527
0.5316	9.0	612	0.5899	0.7986
0.5073	10.0	680	0.6193	0.7491
0.4854	11.0	748	0.5721	0.7845
0.4347	12.0	816	0.6495	0.7562
0.3937	13.0	884	0.6108	0.7597

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1