minang_food_classification

This model was trained from scratch on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.7860
• Accuracy: 0.9278

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: 1e-06
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine_with_restarts
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3423	1.0	45	1.3263	0.7889
1.2638	2.0	90	1.2436	0.8278
1.2055	3.0	135	1.2503	0.8
1.14	4.0	180	1.1486	0.85
1.0908	5.0	225	1.0427	0.8778
1.0258	6.0	270	1.0210	0.8333
0.9776	7.0	315	0.9694	0.8722
0.9306	8.0	360	0.9379	0.8833
0.8985	9.0	405	0.9150	0.8778
0.8624	10.0	450	0.8884	0.8611
0.8243	11.0	495	0.8118	0.9222
0.8017	12.0	540	0.8394	0.8833
0.797	13.0	585	0.7761	0.9056
0.7765	14.0	630	0.7891	0.9111
0.7834	15.0	675	0.7945	0.8889
0.7483	16.0	720	0.7801	0.9
0.74	17.0	765	0.7524	0.9167
0.7315	18.0	810	0.7655	0.9111
0.7468	19.0	855	0.7860	0.8833
0.7393	20.0	900	0.7900	0.9056

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1